org Interactive Fly, Drosophila



Effects of Mutation or Deletion (part 2/2)

Frizzled function in retinal polarity

The photoreceptors within the ommatidia of the Drosophila compound eye form a trapezoid. This occurs in two chiral forms in the dorsal and ventral half of the eye. Ommatidia in the dorsal half of the compound eye are oriented with the R3 photoreceptor cell dorsal and anterior, the R7 photoreceptor being ventral. Ommatidia in the ventral half of the eye are inverted. This asymmetry is established during the progression of the morphogenetic furrow as it moves across the epithelium of the eye imaginal disc from posterior to anterior. As the furrow moves it lays down a new column of ommatidial clusters roughly once every 2 hours. However, the ommatidial clusters in one column are not initiated at the same moment, i.e. the first cluster is formed at the center of the furrow (the midline or future equator); subsequent clusters are formed dorsal and ventral to this at about 10-min intervals. This point at the center of the furrow is known as the firing center, an inductive node which transmits information in two directions, i.e. induction of new ommatidial columns towards the anterior and induction of new ommatidial clusters towards the dorsal and ventral poles (Reifegerste, 1997 and references).

Two manipulations were used to induce ectopic ommatidia, in combination with molecular markers for specific positions in the retinal field. Ectopic furrows were generated by shift of winglessl-12 homozygotes to a nonpermissive temperature for 48 hours. Loss of function patched clones were used to induce ectopic furrows, because patched functions as a negative regulator of furrow initiation. Ectopic morphogenetic furrows induced on the eye field margin (or midline) and those induced in the body of the field have different consequences for the establishment of retinal polarity. Ectopic clones on the midline or margin is associated with ectopic expression of early markers of retinal field polarity, while ectopic expression of clones that do not lie on the margin or midline are not associated with such markers. In cases where clones fail to induce ectopic furrows, such clones can re-specifiy polarity field markers if they lie on the margin or midline. Photoreceptor cells in the ectopic ommatidia formed by patched clones produce axons that do not always follow the normal polarity field toward the posterior and the optic stalk. In cases in which a field of ectopic ommatidial clusters is still disconnected from those formed by the endogenous field, the ectopic clusters do not find a path to the optic stalk, but converge on the center of their local field. This phenomenon may be similar to the development of axon tracts in the insect central nervous system and is consistent with a homophilic axon guidance model (Reifegerste, 1997).

An early equatorial model for retinal polarity is proposed. In this model, early events establish the dorsal/ventral polarity of the retinal field and establish the midline/equator; only later does the furrow initiate and then the firing center follows the midline, but does not form it. This idea is derived from the observation that markers of polarity are expressed in specific parts of the retinal field before furrow initiation. Thus events that initiate furrow movement on the margin or the midline re-specify the field markers, while those that lie off the margin or the midline do not. Evidence for a preexisting field of positional information comes from the characterization of the homeoprotein mirror, which seems to be involved in the establishment of retinal polarity. The gene four-jointed shows a graded expression in equatorial-polar direction along the equator in third instar eye imaginal discs. Four jointed is a putative cell surface or secreted protein. Another candidate for an equatorial signal is Wingless itself. Wingless could act early to signals from the margins inwards. A second signal from the midline could be induced by early Wingless. Mosaic clones for frizzled affect retinal polarity; these have a domineering non-autonomy on adjacent wild type tissue. Proteins similar to Frizzled have been shown to act as Wnt receptors (Reifegerste, 1997 and references).

The position of the future equator is evident in the third instar eye disc prior to the start of ommatidial rotation. New rows emerging from the furrow are initiated at the D/V midline (the site of the future equator) and grow laterally as ommatidia are added to each end in a symmetrical manner. This `center-lateral' growth of a row and ommatidial rotation are both affected in frizzled mutants, suggesting these two events may be linked. However, while the adult phenotypes of fz and strabismus are similar, the center-lateral growth of a row is not altered in stbm mutants. This observation indicates that the two events are genetically separable and that stbm may act downstream of fz, or in a parallel pathway, in its role in orientation but not in its role in the center-lateral growth of new ommatidial rows (Wolff, 1998).

The planar polarity of Drosophila ommatidia is reflected in the mirror-symmetric arrangement of ommatidia relative to the dorso-ventral midline, the equator. This arrangement is generated when ommatidia rotate towards the equator and the photoreceptor R3 displaces R4, creating different chiral forms in each half. Analysis of ommatidia that are mosaic for the tissue polarity gene frizzled (fz) shows that the presence of a single Fz+ photoreceptor cell within the R3/ R4 pair is critical for the direction of rotation and chirality. By analysing clones mutant for seven-up (svp) in which R3/R4 precursors reside in their normal positions and become photoreceptor neurones but fail to adopt the normal R3/R4 fate, it has been found that the R3/R4 photoreceptor subtype specification is a prerequisite for planar polarization in the eye. In mosaic R3/R4 pairs the svp- cell always adopts the R4 position. This bias is reminiscent of what happens in fz mosaic R3/R4 pairs, where the fz- cell also almost always adopts the R4 position. A possible interpretation of the data is that the svp mutant cell is not able to receive the polarity signal or to interpret it (or to communicate with the other cell of the R3/R4 pair). Mutations in the rough gene cause the mis-specification of R2 and R5 and their transformation to an R3/4 pair as seen by their expression of svp and their dependence on svp to develop as outer photoreceptors. In genotypes where too many cells adopt the R3/R4 fate, ommatidial polarity is also disturbed. This defect could arise because, in a situation with too many R3/R4 cells within a cluster, there is crosstalk/competition for the R3 fate between more than two cells that confuses the cluster as a whole. Taken together, these data imply that correct specification of a single R3 cell per ommatidium is a prerequisite for the normal interpretation of the Fz-mediated polarity signal (Fanto, 1998).

Planar polarity is seen in epidermally derived structures throughout the animal kingdom. In the Drosophila eye, planar polarity is reflected in the mirror-symmetric arrangement of ommatidia (eye units) across the dorsoventral midline or equator; ommatidia on the dorsal and ventral sides of the equator exhibit opposite chirality. Photoreceptors R3 and R4 are essential in the establishment of the polarity of ommatidia. The R3 cell is thought to receive the polarizing signal, eminating from the equator, through the receptor Frizzled (Fz), before or at higher levels than the R4 cell, generating a difference between neighbouring R3 and R4 cells. Both loss-of-function and overexpression of Fz in the R3/R4 pair result in polarity defects and loss of mirror-image symmetry. Notch and Delta (Dl) are identified as dominant enhancers of the phenotypes produced by overexpression of fz and dishevelled (dsh); dsh encodes a signaling component downstream of Fz, and it is shown that Dl-mediated activation of Notch is required for establishment of ommatidial polarity. Whereas fz signaling is required to specify R3, Notch signaling induces the R4 fate. These data indicate that Dl is a transcriptional target of Fz/Dsh signaling in R3, and Dl activates Notch in the neighboring R4 precursor. This two-tiered mechanism explains how small differences in the level and/or timing of Fz activation reliably generate a binary cell-fate decision, leading to specification of R3 and R4 and ommatidial chirality. How Notch signaling induces the R4 fate remains unclear, as it usually represses photoreceptor development at this stage. However, the precursor cells are already committed to form the R3/R4 pair by transcription factors (such as Seven-up) that are required for both R3/R4-cell fate and polarity generation (Fanto, 1999).

The Drosophila eye, a paradigm for epithelial organization, is highly polarized with mirror-image symmetry about the equator. The R3 and R4 photoreceptors in each ommatidium are vital in this polarity; they adopt asymmetrical positions in adult ommatidia and are the site of action for several essential genes. Two such genes are frizzled (fz) and dishevelled (dsh), the products of which are components of a signaling pathway required in R3, and which are thought to be activated by a diffusible signal. The transmembrane receptor Notch is required downstream of dsh in R3/R4 for them to adopt distinct fates. By using an enhancer for the Notch target gene Enhancer of split mdelta, it is shown that Notch becomes activated specifically in R4. Analyzing the regulation of E(spl)mdelta, it has been found that this target of Notch is expressed specifically in R4. Transiently reducing Notch activity for 6 hours in late-third-instar larvae, using temperature sensitive Notch, leads to a loss of E(spl)mdelta expression, whereas transient activation of a constitutively active Notch derivitive has the converse effect. Genetic experiments show the importance of Dsh in the establishment of eye polarity. The mutant protein coded for in dishevelled mutants has impaired signaling, but the phenotype can be partially rescued by overexpressing downstream components. When E(spl) proteins are expressed ectopically in dsh mutant discs, the eyes are less roughened and more ommatidia have correct rotation and chirality, compared to when E(spl) proteins are not ectopically expressed. It is proposed that Fz/Dsh promotes expression of the Notch ligand Delta and inhibits Notch receptor activity in R3, creating a difference in Notch signaling capacity between R3 and R4. Subsequent feedback in the Notch pathway ensures that this difference becomes amplified. This interplay between Fz/Dsh and Notch indicates that polarity is established through local comparisons between two cells and explains how a signal from one position (for example, the equator in the eye) could be interpreted by all ommatidia in the field. Additional targets of Notch needed to specify R4 identity may include strabismus. Ommatidial polarity also requires spiney-legs (sple). Sple is normally inhibitory to Notch in R3, because high levels of E(spl)mdelta expression are found in sple mutants. The inhibitory effect of Sple provides a second mechanism for polarizing Notch signaling in R3/R4 that could be coordingated by Fz/Dsh (Cooper, 1999).

The pattern of connections between R1-R6 neurons and their targets in the lamina is one of the most extraordinary examples of connection specificity known. An interwoven set of connections precisely maps R cells in different ommatidia (that 'see' the same point in space) onto the same group of postsynaptic cells, the lamina cartridge. R1-R6 cells that see the same point in space are distributed over six neighboring ommatidia as a consequence of the curvature of the eye and the angular placement of their light-sensing organelles. Conversely, each of the R1-R6 axons from a single ommatidium sees a different point in space and connects to a different set of lamina target neurons arranged in an invariant pattern. Each cartridge is innervated by a complete set of R1-R6 neurons from six different ommatidia (i.e., an R1 from one ommatidium, an R2 from another, and so on). By superimposing multiple inputs from the same point in visual space upon a single synaptic unit, the signal-to-noise ratio of the response to a signal in the visual field is enhanced. This phenomenon is called neural superposition (Clandinin, 2000 and references therein).

The R1-R6 projection pattern develops in two temporally distinct stages. During the third larval stage, R cells extend axons into the brain, where they terminate between two layers of glia, forming the lamina plexus. These glia act as intermediate targets for R1-R6 neurons. R cell axons induce the differentiation and organization of lamina target neurons and glia. At this stage of development, R cell axons from the same ommatidium form a single fascicle. A column of lamina neurons forms above the lamina plexus, in tight association with a single R cell axon fascicle. By the sequential addition of ommatidial bundles and their associated columns of lamina neurons, a precise retinotopic map forms in which fascicles from neighboring ommatidia terminate adjacent to each other. As lamina neurons differentiate, they send axons along the surface of R cell axons through the plexus and fasciculate with R7 and R8 as they project into the medulla. Although lamina neurons are in close association with R cell axons at this early stage, no synaptic contacts are formed (Clandinin, 2000).

In the second phase of development, ~30 hr after reaching the lamina plexus, R cell axons defasciculate from each ommatidial bundle and project across the surface of the lamina to their synaptic partners, making the pattern of connections characteristic of neural superposition. Growth of R cell axons toward their targets occurs approximately simultaneously in all ommatidial bundles and is presaged by an invariant sequence of contacts between R cell growth cones. This reorganization of terminals converts a strictly anatomical retinotopic map that reflects neighbor relationships between ommatidia into a new topographic map that reflects R cell visual response and reconstructs visual space in the first layer of the optic ganglion (Clandinin, 2000 and references therein).

R cell projections from a single ommatidium display two prominent features. (1) Each R cell axon terminates in an invariant position relative to the other axons from the same ommatidial fascicle. (2) The projection is oriented with respect to the dorsoventral midline of the eye (i.e., the equator), with the R3 axon extending toward the equator -- as a result, the projection patterns on opposite sides of the dorsoventral midline of the eye are mirror images. Using mutations that eliminate specific subsets of R cells or alter ommatidial polarity, tests were performed to see whether R cell synaptic specificity requires interactions among neighboring afferent axons or reflects independent navigation of each axon to its target. It has been demonstrated that interactions between specific R cells are required for target selection, and it is proposed that the precise composition of R cell axons within a fascicle plays a critical role in target specificity (Clandinin, 2000).

Neural superposition was first noted 90 years ago and the R1-R6 connection pattern in the lamina was first described using serial reconstruction of electron microscopic images in 1965. This pattern is cited as a classic example of extreme connection specificity. However, mechanistic analysis of this pattern was prevented by the absence of a rapid method for assessing R cell projections. In particular, the complexity of the pattern precludes conventional approaches based on visualizing all R cell axons in the target region, yet the assessment of connection specificity requires visualization of all R cell axons from one ommatidium. A method has been developed to label individual ommatidia with DiI and visualize the projection pattern using confocal microscopy. R1-R6 axons form a single bundle as they project into the brain. They defasciculate, project across the surface of the lamina, and then turn 90° and extend into the lamina cartridge. R cell axons elaborate a complex en passant presynaptic structure with lamina interneurons within the lamina cartridge. The axons of R7 and R8 project through the lamina, into the medulla. The relative positions of lamina targets chosen by each R1-R6 growth cone are invariant between ommatidia. This labeling method facilitates analysis of R1-R6 projections in various genetic backgrounds and creates a unique experimental system in which synaptic partner choices made by identified neurons can be directly assessed (Clandinin, 2000).

Serial electron microscopic reconstruction studies have revealed that, during pupal development, individual R cell axons leave their original bundle and migrate outward, in the precise direction of their final targets. This process was visualized using confocal microscopy. Early in pupal development, each ommatidial bundle forms a compact mass of expanded growth cones in the lamina plexus. This spherical mass then flattens, as distinct filopodial extensions corresponding to individual R cell axons become visible. This pattern of connections forms within a spatially patterned environment containing lamina target neurons and glial cells, as well as R cell axons. Since extension from the bundle is not preceded by extensive filopodial exploration, interactions between axons within ommatidial bundles may specify the initial trajectory of each growth cone. To address whether cell intrinsic mechanisms or interactions between R cell growth cones or both control target specificity, R cell projections were examined in mutant animals lacking specific subsets of R1-R6 cells. R cell axons from single ommatidia were labeled with DiI and visualized by confocal microscopy. In this series of experiments, animals were analyzed in which the eye was genetically mutant and the lamina neurons and glia in the target were wild type. Three mutant backgrounds were examined: (1) phyllopod, in which R1, R6, and R7 are transformed into nonneuronal cone cells; (2) lozengesprite, in which R3 and R4 are transformed into R7 cells; and (3) seven-up, in which R1, R3, R4, and R6 are transformed into R7 cells (Clandinin, 2000).

The first step of lamina target innervation is the coordinated defasciculation of R cell axons from bundles comprising axons from the same ommatidium. To determine whether interactions between specific subsets of R1-R6 axons are necessary for this defasciculation, R cell projections were assessed in phyllopod, seven-up, and lozengesprite mutants. In all three of the R cell transformation mutants examined, R cell axons migrated outward from the bundle. In particular, 4 R cell fibers in the lamina of 14/15 phyllopod mutant animals (missing R1, R6, and R7) and 20/24 lozengesprite mutants (missing R3 and R4) defasciculated from the bundle and projected to local targets. Similarly, in 17/23 seven-up mutants (missing R1, R3, R4, and R6), it was observed that the two remaining R cell axons defasciculated from the ommatidial bundle and innervated separate cartridges. In some cases, additional R cell axons also defasciculated, consistent with the reported incomplete expressivity of cell fate transformations in these mutants. In each case, axons projected to lamina targets in the local environment of the fascicle terminus. It is concluded that each R cell subtype is programmed to initiate a search for targets in a local region of the lamina target, independent of interactions between other R cell subtypes. In the following sections, whether interactions between specific R1-R6 cells regulate target specificity is assessed (Clandinin, 2000).

Two models could explain the mechanisms that determine the precise projection of R3 and R4 axons toward the dorsoventral midline and, by extension, the relative orientations of the other R cell axons. The growth cones of R3 and R4 may respond to an orienting cue in the lamina that promotes extension toward the dorsoventral midline. Alternatively, the orientation of R cell bodies in the retina may determine the orientation of R cell growth cones in the lamina, independent of any environmental cues. To assess the role of ommatidial polarity on projection specificity, projections from misoriented ommatidia were assessed (Clandinin, 2000).

If a lamina cue can promote equatorial extension of the R3 and R4 axons, ommatidia that rotate incorrectly should project their axons normally, toward the equator. Alternatively, if ommatidial orientation determines the direction of axon projection in the lamina, incorrectly oriented ommatidia should project their R3 and R4 axons away from the equator (Clandinin, 2000).

In wild-type animals, ommatidia are mirror image reflected about the dorsoventral equator of the eye. R cell projections are also mirror image symmetric about the equator but are rotated 180° with respect to the retina. That is, while the R3 cell body is oriented toward the pole in each ommatidium, its axon projects toward the equator in the lamina. This rotation is generated by a twist in the axon fascicle that occurs between the retina and the lamina (Clandinin, 2000).

To test the effects of large changes in ommatidial orientation, two mutations, spiny legs (in homozygous animals) and frizzled (in somatic mosaic animals in which a mutant eye projects to a wild-type target), were examined. In these mutants, ommatidia frequently adopt orientations that are 180° rotated; that is, the R3 cell body is frequently oriented toward the equator in the eye. In these two mutant backgrounds, the orientation of projections from ommatidia that were correctly oriented was normal. Therefore, neither gene is required for R cell axons to respond to orienting cues in the target. However, almost 90% of the ommatidia that were ~180° misoriented in the eye made projections that were also 180° misoriented in the lamina. Rare, abnormal projections of single R cell axons in both of these mutant backgrounds were observed, irrespective of ommatidial orientation. Therefore, the orientation of R cell projections along the dorsoventral axis of the lamina is largely determined by the orientation of ommatidia in the retina (Clandinin, 2000).

Three exceptional cases, in which misoriented ommatidia projected axons toward the equator, were observed. Thus, a cue in the lamina may reinforce the ommatidial orientation cue to ensure the correct direction of outgrowth along the dorsoventral axis. To test whether such a cue contributes to directionality of R cell projections, a mutation that causes a more moderate defect in ommatidial orientation was examined. In nemo mutant animals, ommatidia are misoriented up to 45°. If ommatidial orientation directly determines the directionality of R cell projections, they would be misoriented 45° with respect to the equator; the angle between ommatidial orientation and the axon projection pattern would remain 180°. However, while ommatidial orientation was disrupted in nemo, R cell projections were normal with respect to the equator. This observation suggests that in addition to ommatidial polarity, a cue in the lamina can influence R cell projection orientation (Clandinin, 2000).

It is concluded that interactions between R cell afferents play a crucial role in target specification, and it is proposed that the spatial relationships between axons within a fascicle influence synaptic specificity. It is hypothesized that the interactions between R cell subtypes that are required for target specificity are mediated by direct contacts between specific growth cones. R3 and R4 are required for the remaining R cell axons to choose their normal targets. R1 and R6 are required for R2 and R5 projections but are not required for the projections of R3 and R4. These interactions could occur between growth cones from the same or neighboring ommatidial bundles. The characteristic morphological changes of these growth cones as revealed through electron microscopic reconstruction studies are consistent with the notion that precise spatial relationships between specific growth cones within the lamina plexus are required for these critical interactions to occur. This sequence of interactions determines the relative positions of targets chosen by R cell axons from the same ommatidium (Clandinin, 2000).

R cell transformation mutants could disrupt these interactions in two ways. First, transformation of specific R cells could directly disrupt the instructive signals between R cell growth cones within the plexus that determine growth cone trajectories. Alternatively, these mutations could affect the interactions indirectly, by disrupting the spatial relationships between the remaining R cell axons. That is, outgrowth trajectory could be determined passively by the position each growth cone occupies as it leaves the ommatidial fascicle. In this view, these mutant backgrounds alter the composition of axons within each ommatidial bundle and, hence, disrupt the precise packing of axons within the fascicle. The differential requirements for particular R cell subtypes would reflect their specific roles in directing the spatial relationships between growth cones within the fascicle, rather than interactions between specific growth cones in the target region (Clandinin, 2000).

Ommatidial polarity is defined by the relative positions of R cells within an ommatidium. Each R cell occupies an invariant position; R1-R6 cells within each ommatidium create a pattern that is mirror-image symmetric about the dorsoventral midline of the eye. The observation that ommatidial polarity determines projection orientation requires that the spatial relationships between R cell bodies be maintained in ommatidial axon fascicles. Indeed, a striking feature of Drosophila visual system connectivity is the perfect conservation of spatial relationships between R cell axons, both within each bundle and with respect to the dorsoventral axis of the eye. It is hypothesized that the developmental mechanisms that determine where each R cell differentiates in the retina also control where each R cell axon lies within the fascicle and how the fascicle is oriented along the dorsoventral axis (Clandinin, 2000).

Control of projection orientation by ommatidial polarity also requires that the relative positions of R cell axons within a fascicle, as well as the dorsoventral orientation of the fascicle itself, be 'read out' in the lamina. In this view, the relative positions of axons within the fascicle allows the specific interactions between growth cones that control synaptic specificity to 'self-organize' the pattern of targets. Since axons from both correctly oriented and misoriented ommatidia choose targets arranged in a normal pattern, these interactions between growth cones must occur independent of orientation along the dorsoventral axis. In this model, fascicle orientation determines whether the pattern of targets chosen is oriented either dorsally or ventrally but does not determine the relative positions of the targets within the pattern. This approach of 'encoding' the spatial arrangement of sensory neuron cell bodies within an axon fascicle followed by 'reading out' the preserved orientation cues within the target may provide a general mechanism to generate highly precise patterns of connections (Clandinin, 2000).

Projection specificity in the lamina is not solely controlled by interactions between R cell axons. The observation that the R cell projections are correctly oriented in nemo mutant animals provides evidence that a cue(s) in the target can reorient R cell axons. Such a cue need only orient a subset of R cell axons, likely R3 and R4; these axons could then organize the remaining R1-R6 projections. This cue could be a weak signal that directs R3 and R4 axon outgrowth toward the equator. Alternatively, this cue could simply confine the outgrowth of R3 and R4 to the dorsoventral axis, without determining whether outgrowth is either dorsal or ventral. These target-derived cues could correct for small variations in ommatidial orientation (and fascicle orientation) that exist in the pupal eye prior to programmed cell death in the retina (Clandinin, 2000).

Previous studies suggest that connection specificity can be generated by activity-dependent refinement of synaptic contacts. Indeed, interactions between synapses mediated by electrical activity are clearly important for several aspects of neuronal connectivity in the vertebrate visual system, including the maintenance of ocular dominance columns and formation of eye-specific layers in the lateral geniculate nucleus. In the Drosophila retina-lamina projection, the role of neuronal activity is unclear. These connections likely develop independent of visual input since the projections form when R cells display little light-evoked response. Moreover, phototransduction-defective mutants display normal numbers of R cell termini in each cartridge, suggesting that R cell target selection occurs normally in these backgrounds. These connections also form prior to the development of morphologically distinct synaptic contacts between R cells and lamina neurons (Clandinin, 2000).

To test the effects of changes in neuronal activity and synaptic transmission on R cell targeting, known mutations affecting both sodium and potassium channel subunits, as well as synaptotagmin were examined. None of these backgrounds displays defects in R cell synaptic specificity. However, these genetic approaches are confounded by the significant molecular redundancy present in these protein families within the Drosophila genome. Synaptobrevin-mediated vesicle release was disrupted by expressing tetanus toxin specifically in R cells. Synaptobrevin blockade, however, affects expression of cell adhesion molecules in the Drosophila visual system and disrupts axonal morphology. Hence, it remains unclear whether disruption reflects a role for synaptic transmission in R cell target specificity or is an indirect result of effects on other cellular processes. In summary, while these experiments demonstrate that developmental mechanisms that are likely to be independent of neuronal activity are sufficient to generate the precise pattern of retina-lamina connections, a role for neuronal activity cannot be excluded (Clandinin, 2000).

The cellular mechanisms described here provide a conceptual framework for understanding the molecular basis of synaptic specificity. While the DiI method facilitates the analysis of R1-R6 specificity on a scale sufficient to analyze many mutants, it is too laborious to accommodate large-scale screening. Hence, a genetic screen based on visual behavior driven specifically by R1-R6 is required to extend these studies to the molecular level. A wealth of visual behaviors have been described in Drosophila, one of which, the optomoter response, is mediated by these cells. Techniques that generate mosaic flies in which only R cells are made homozygous for randomly induced mutations, while the rest of the fly is heterozygous, have recently been described. Currently, projects are underway, combining this specific behavioral screen with genetic mosaics, in order to screen for genes controlling R1-R6 synaptic specificity (Clandinin, 2000).

The ommatidia of the Drosophila eye initiate development by stepwise recruitment of photoreceptors into symmetric ommatidial clusters. As they mature, the clusters become asymmetric, adopting opposite chirality on either side of the dorsoventral midline and rotating exactly 90°. The choice of chirality is governed by higher activity of the frizzled (fz) gene in one cell of the R3/R4 photoreceptor pair and by Notch-Delta (N-Dl) signaling . The 90° rotation also requires activity of planar polarity genes such as fz as well as the roulette (rlt) locus. Two regulators of EGF signaling, argos and sprouty (sty), and a gain-of-function Ras85D allele, interact genetically with fz in ommatidial polarity. Furthermore, argos is required for ommatidial rotation, but not chirality, and rlt is a novel allele of argos. Evidence is presented that there are two pathways by which EGF signaling affects ommatidial rotation. In the first, typified by the rlt phenotype, there is partial transformation of the 'mystery cells' toward a neuronal fate. Although most of these mystery cells subsequently fail to develop as neurons, their partial transformation results in inappropriate subcellular localization of the Fz receptor, a likely cue for regulating ommatidial rotation. In the second, reducing EGF signaling can specifically affect ommatidial rotation without showing transformation of the mystery cells or defects in polarity protein localization (Strutt, 2003).

Mutations in fz result in defects in planar polarity of the eye, characterized by ommatidia taking on random chirality, or no chirality, and rotating randomly. A hypomorphic combination of fz alleles fz19/fz20 results in a weak eye phenotype in which only 9% of ommatidia show polarity defects. This phenotype is strongly enhanced by removing one copy of the dishevelled (dsh) gene, which acts downstream of Fz in polarity signaling (Strutt, 2003).

In order to identify additional factors involved in regulating ommatidial polarity, a large-scale genetic screen was carried out for loci interacting with fz. Unexpectedly, the principle factors identified were components of the EGF signaling pathway: three complementation groups corresponded to the genes argos, sty, and Ras85D. argos encodes an inhibitory ligand for the Drosophila EGF receptor. The new allele isolated in this screen (argos5F4) and two independent alleles enhanced the fz19/fz20 phenotype, such that about 20% of ommatidia had polarity defects. Similarly, the fz19/fz20 phenotype was also enhanced by two novel alleles and three known alleles of sty, which encodes a cytoplasmic protein that inhibits the Ras signaling pathway. Finally, the 2F4 enhancer mutation had an unusual dominant phenotype, in which a small number of ommatidia had extra R7 cells and very rare defects in specification of outer photoreceptors; also, extra vein tissue was seen in the wing. This phenotype is reminiscent of dominant mutations in the MAPK gene rl (rlSem, and the extra R7 cell phenotype is increased by removing one copy of the negative Ras pathway components sty, Gap1, and yan. Transheterozygotes of 2F4 and loss-of-function Ras85D mutations result in a weak Ras85D phenotype, in which outer photoreceptors were lost from many ommatidia. This phenotype suggests that 2F4 might be a Ras85D allele. This was confirmed by sequencing of the Ras85D gene in 2F4 mutants, which revealed a mutation of Ala59 to Thr. Interestingly, this mutation is a weak activating mutation found in viral oncogenes. Hence, mutations in three EGF pathway components, each of which are predicted to increase levels of pathway activity, are dominant enhancers of an fz ommatidial polarity phenotype (Strutt, 2003).

sty mutants have a severe rough eye phenotype characterized by transformation of cone cells to R7 photoreceptors and, less frequently, of mystery cells into outer photoreceptors. This phenotype is sufficiently strong that it is not possible to deduce from adult eye sections whether the ommatidia are also misrotated. However, examination of eye imaginal discs from sty homozygotes shows that the developing ommatidial clusters are not uniformly rotated relative to each other (Strutt, 2003).

It is concluded that EGF signaling is required for correct ommatidial rotation. A fz ommatidial polarity phenotype is dominantly enhanced by argos, Ras85D2F4, and sty, all of which result in excess EGF pathway activation. Additionally, ommatidial rotation defects are seen in conditions in which EGF pathway activity is either increased or decreased (Strutt, 2003).

It is proposed that there are two mechanisms by which EGF signaling affects ommatidial rotation. The first is that this is a result of mystery cells inappropriately taking on an R3/R4 fate. In argosrlt, most ommatidia show partial transformation of mystery cells into R3/R4 photoreceptors. Although most of these extra R3/R4 cells do not ultimately differentiate into neurons, Fz-GFP is mislocalized in them at the time of ommatidial rotation. Since fz is required in the R3/R4 photoreceptor pair for correct ommatidial chirality and rotation, the presence of extra cells containing localized Fz-GFP could be providing the ommatidium with conflicting cues that disrupt normal rotation (Strutt, 2003).

Frizzled function and Fat

The fat gene negatively controls cell proliferation in a cell autonomous manner. The Fat protein (with 5,147 amino acids) contains four major regions. Beginning by the N-terminus there are 34 cadherin-like domains, five EGF-like repeats interspersed with two laminin A-G chain motifs, a transmembrane domain and a novel cytoplasmic domain (Mahoney, 1991). Several cell behavior parameters of mutant alleles of fat ( ft) have been studied in Drosophila imaginal wing disc development. Mutant imaginal discs continue growing in larvae delayed in pupariation and can reach sizes of several times those of wild-type. Their growth is, however, basically allometric. Homozygous ft cells grow faster than their twin cells in clones and generate larger territories, albeit delimited by normal clonal restrictions. Moreover, ft cells in clones tend to grow towards the wing proximal regions. These behaviors can be related with failures in cell adhesiveness and cell recognition (Garoia, 2000).

Fat also plays an important role in planar polarity. This phenomenon is evidenced by the coordinated orientation of ommatidia in the Drosophila eye. Planar polarity requires that the R3 photoreceptor precursor of each ommatidium has a higher level of Frizzled signaling than its neighboring R4 precursor. Two cadherin superfamily members, Fat and Dachsous, and the transmembrane/secreted protein Four-jointed play important roles in this process. The data support a model in which the bias of Frizzled signaling between the R3/R4 precursors results from higher Fat function in the precursor cell closer to the equator -- the cell that becomes R3. Evidence is also provided that positional information regulating Fat action is provided by graded expression of Dachsous across the eye and the action of Four-jointed, which is expressed in an opposing expression gradient and appears to modulate Dachsous function. It is suggested that the presence of relatively higher Ds function in the polar cell could result in a difference in Ft function between the R3/R4 precursors by either inhibiting Ft function in a cell-autonomous fashion or by stimulating Ft function in the equatorial cell. The difference in Ft function between the precursor cells biases Fz signaling so that the equatorial cell has higher Fz activity (Yang, 2002).

The first indication that Ft functions during PCP signaling in the eye came from examining flies homozygous for the weak, viable ft allele, ft1. A small fraction of ommatidia with reversed dorsal-ventral (d-v) polarity were consistently observed. To examine the effects of stronger ft alleles on ommatidial polarity, FRT-mediated mitotic recombination was used to generate clones of cells homozygous for the lethal ft alleles, ftl(2)fd or ftGr-V. Ommatidia located within the ft mutant tissue were constructed normally, but they frequently adopted the reversed d-v polarity form (~40% for ftl(2)fd, ~50% for ftGr-V). In addition to polarity reversals within the ft mutant tissues, occasional reversals of polarity in wild-type ommatidia bordering the polar side of the mutant tissue were also observed. This effect resembles the nonautonomous phenotypes previously reported for fz mutant clones in the eye (Yang, 2002).

The presence of a randomized pattern of d and v type ommatidia within ft mutant tissue suggests that Ft is required to correctly bias R3/R4 specification. To confirm that the initial pattern of R3 and R4 specification is randomized in ft mutant ommatidia, the expression pattern of an R4-specific marker, E(spl)mdelta0.5, was examined in ftl(2)fd and ftGr-V clones. This marker consists of a portion of the enhancer region of the E(spl) gene, a transcriptional target of N activation, fused to a ß-galactosidase (ß-gal) reporter. In wild-type ommatidia, this marker is initially expressed in both R3/R4 precursor cells, but then becomes stronger in the polar cell since this cell is specified as R4. Within the ft ommatidial clusters, the R4-specific marker is still strongly expressed in only one member of the R3/R4 pair. However, the cell expressing the R4 marker frequently occupies the position normally taken by the equatorial cell, indicating that the pattern of R3/R4 cell fate specification is reversed (Yang, 2002).

Because the presence of higher Fz signaling in the equatorial member of the R3/R4 precursor pair is a crucial determinant of R3/R4 specification, these results suggest that Ft may function in the placement or interpretation of positional cues controlling the equatorial/polar bias of Fz signaling. This model predicts that while Fz signaling should still occur in the absence of Ft, the pattern of Fz activation within an R3/R4 pair should be randomized with respect to the equator. Consistent with this prediction, two important differences between the ft and fz mutant phenotypes indicate that Fz signaling remains intact in ft ommatidia. (1) Previous studies have shown that many fz ommatidia are incorrectly formed and fail to have distinctly specified R3 and R4 cells. This phenotype indicates that the processes preventing both R3 and R4 precursor cells from adopting the same fate depend on Fz function and are inefficient when only Dl/N-mediated lateral inhibition is used to specify R3 and R4. (2) fz ommatidia frequently either fail to rotate or rotate incorrectly even when R3 and R4 cells are specified. In contrast, most ft ommatida contain uniquely specified R3 and R4 cells and rotate in the proper direction for their pattern of R3/R4 specification. These differences between the fz and ft phenotypes suggest that Fz signaling remains functional in the absence of Ft function (Yang, 2002).

To show that the absence of Ft function causes Fz signaling to occur in a randomly biased pattern within R3/R4 precursor pairs, eye discs containing ft mutant clones were immunostained for Flamingo (Fmi)/Starry Night (Stan) protein. Previous work has suggested that Fmi and Fz function together in a signaling complex at the proximal/distal (p/d) junctions between wing cells. Furthermore, the accumulation of Fmi at a p/d cell-cell boundary depends on the presence of differences in Fz signaling levels between the two cells. To first show that Fmi functions during PCP signaling in the eye, clones of cells homozygous for a strong loss-of-function fmi allele (fmiE59) were examined. Dramatic polarity defects were found that resembled those seen in fz ommatidia. The defects included aberrant rotation and a lack of distinct R3/R4 fates (Yang, 2002).

The frizzled (fz) gene of Drosophila is required for planar polarity establishment in the adult cuticle, acting both cell autonomously and nonautonomously. These two activities of fz in planar polarity are temporally separable in both the eye and wing. The nonautonomous function is dishevelled (dsh) independent, and its loss results in polarity phenotypes that resemble those seen for mutations in dachsous (ds). Genetic interactions and epistasis analysis suggest that fz, ds, and fat (ft) act together in the long-range propagation of polarity signals in the eye and wing. Evidence has been found that polarity information may be propagated by modulation of the binding affinities of the cadherins encoded by the ds and ft loci (H. Strutt, 2002).

In the wild-type wing, each cell produces a single trichome at its distal vertex that points distally. Flies that lack fz function exhibit defects in trichome polarity. Most cells still produce a single trichome, but these are arranged in a distinctive swirling pattern known as the fz/in-like pattern. Furthermore, in the pupal wing, there is no asymmetric localization of the polarity proteins Fz, Dsh, Fmi, Pk, and Dgo and the majority of trichomes form in the cell center. An almost identical phenotype is caused by mutations that remove only the cell-autonomous functions of fz or mutations in the cell autonomously-acting polarity genes dsh and fmi. This indicates that this phenotype is likely to be solely the result of removing cell-autonomous polarity gene function (H. Strutt, 2002).

In contrast, ds is thought to be required only for the nonautonomous transmission of polarity information. Wings from ds individuals show a trichome swirling pattern, which is distinct from the fz/in pattern. Asymmetric polarity protein localization still occurs—albeit often in an aberrant pattern Similarly, cells adjacent to a fz clone, which have aberrant polarity because of the nonautonomous phenotype of fz, also asymmetrically localize polarity proteins (H. Strutt, 2002).

On the basis of these observations, it seems likely that the loss of fz function from the entire wing results in both cell-autonomous and -nonautonomous phenotypes, but that the former masks the latter. To study the phenotype caused by loss of the nonautonomous component only, a method of rescuing only the cell-autonomous phenotype must be devised. It was hypothesized that the nonautonomous function of fz might precede the cell-autonomous function. If this is correct, it should be possible to rescue only the cell-autonomous function of fz by adding back fz activity to a fz mutant fly at a stage of development after the nonautonomous requirement, but before the autonomous requirement. To test this, a transgene was used consisting of the Actin5C promoter separated from the fz coding sequence by an FRT-flanked transcription termination sequence. The fz coding sequence is fused to the coding sequence of GFP, to permit monitoring of fz expression, giving rise to the transgene Actin>stop>Fz-GFP. Expression of Fz-GFP from this transgene can be induced by heat-shocking flies that also carry a transgene containing yeast FLP recombinase under control of the hsp70 promoter (H. Strutt, 2002).

This expression system was used to rescue the phenotype of strong fz mutations that lack both cell-autonomous and -nonautonomous activity. Activation of Fz-GFP expression at time points up to about 6 hr after prepupa formation (APF) results in a wild-type trichome polarity pattern. However, activation of Fz-GFP expression at about 6 hr APF results in a weak polarity defect in some parts of the wing. Successively later activation results in an increasingly strong trichome polarity phenotype. Remarkably, this trichome polarity phenotype appears more ds-like than fz/in-like, in particular, showing complete inversions of trichome polarity in some proximal regions of the wing. Activation of Fz-GFP expression at 16 hr APF produces a pattern most similar to that seen in strong ds mutants. Activation at 24 hr APF gives a phenotype that is interpreted as a stronger form of the reported ds-like pattern (and that is stronger than the fz/in pattern). Notably, the characteristic wing shape and vein phenotypes seen in ds mutations are not observed, suggesting that the effects of ds on trichome polarity and wing morphology could be separable functions (H. Strutt, 2002).

Activation of Fz-GFP expression at later time points results in no increase in strength of the ds-like polarity phenotype. Instead, after 27 hr APF, progressively weaker phenotypes are seen that resemble a mixture of the ds-like and fz/in-like patterns, until about 31 hr APF, when the adult polarity pattern is typically fz/in-like and is identical to that produced in the absence of Fz-GFP expression (H. Strutt, 2002).

The effects of expression of Fz-GFP at different time points were also examined in the pupal wing at the time of trichome initiation (about 32 hr APF), monitoring Fz-GFP localization and trichome formation. Activation of Fz-GFP expression up to about 24 hr APF results in asymmetric Fz-GFP localization in an aberrant pattern and trichome formation at the corresponding cell edge. Activation at time points from 24 hr APF onward results in reduced asymmetric protein localization and an increase in the number of trichomes initiating away from the cell edge, until, by 30 hr APF, little Fz-GFP expression is seen at the time of trichome initiation, and trichomes are forming at the cell center (H. Strutt, 2002).

Thus, the temporal requirement for fz shows two phases. Activation of Fz-GFP expression prior to 6 hr APF elicits complete rescue of fz functions. Between 6 and 24 hr APF, lack of fz activity results in a ds-like polarity pattern (which gets progressively stronger, the longer fz activity is not present), but asymmetric polarity protein complexes still form, and trichomes emerge at the corresponding sites at the cell periphery. Between 24 and 31 hr APF, loss of fz activity results in a progressively more dominant fz/in-like trichome polarity pattern, failure to form asymmetric polarity protein complexes, and formation of trichomes in the cell center (H. Strutt, 2002).

Consistent with the hypothesis that the early phase of fz activity is required for nonautonomous polarity signaling, expression of Fz-GFP at 12 hr APF largely rescues the phenotype of an fz allele that is classified as lacking only cell-autonomous activity. At first sight, it is surprising that such an allele is not completely rescued. However, it has recently been reported that such alleles are also partially deficient in nonautonomous signaling activity (H. Strutt, 2002).

The results show that there are two temporally separable activities of fz. The later corresponds to the well-characterized fz cell-autonomous function. The early fz activity, the loss of which produces a more ds-like phenotype, is the fz nonautonomous pathway. Notably, this early activity is required over a significant period of time (16-18 hr in the eye and wing). In addition, this nonautonomous activity is dsh independent in the eye (H. Strutt, 2002).

A number of conclusions follow from these observations: (1) fz exhibits a similar dsh-independent nonautonomous activity in two different tissues (eye and wing), and, therefore, this is likely to be a conserved pathway; (2) the similarity of the fz nonautonomous phenotype to the ds phenotype and genetic interactions between these loci suggests that these molecules might cooperate in a common mechanism to propagate polarity signals; (3) since the nonautonomous activity of fz precedes the autonomous activity, the former is apparently not dependent on the latter. This supports models of polarity patterning, in which a long-range signal is propagated through the tissue prior to the cell-autonomous response to that signal (H. Strutt, 2002).

It is noted that, although these two activities of fz are temporally separable, during normal development it is likely that the period of nonautonomous activity nevertheless overlaps the beginning of autonomous polarity functions (H. Strutt, 2002).

In the eye, the clonal phenotypes of long-range patterning factors, such as the canonical Wnt pathway, JAK/STAT, and fj, have led to models in which they act to establish an activity gradient of a polarity signal (the 'secondary signal'). There are a number of reasons for supposing that these three pathways perform distinct functions that differ from that of the fz nonautonomous activity and that fz is likely to act downstream of these other factors: (1) gradients of components of all three of these pathways are apparent in the second instar stage of development, whereas fz nonautonomous activity in the eye is required over a period of up to 16 hr in the third instar; (2) fz nonautonomous activity is epistatic to (functions downstream of) dsh nonautonomous activity, and there are no genetic interactions between fz nonautonomous activity and canonical Wnt, JAK/STAT, or fj activities; (3) all three activities show similar nonautonomous clonal phenotypes with normal ommatidial polarities in the center of clones, but fz exhibits partial randomization of ommatidial polarities inside the clones (H. Strutt, 2002).

Conversely, there are a number of reasons for thinking that fz nonautonomous activity in the eye is closely related to ds and ft function. The phenotypes of clones lacking early fz function are similar to those of ds clones and ft clones. Furthermore, there are strong genetic interactions between these factors. Finally, an epistasis test between the clonal phenotypes of fz and ds gives an apparently additive (or possibly synergistic) phenotype. These results are consistent with fz acting jointly with ds and ft in the nonautonomous propagation of polarity information. A similar function for ds has been suggested on the basis of studies in the wing, it having been shown that ds nonautonomously affects trichome polarity and that it is likely to be involved in the maintenance or propagation of an fz-dependent nonautonomous polarity signal (H. Strutt, 2002).

Thus, overall data from both the eye and wing support fj acting upstream of ds and ft, which then act jointly with fz nonautonomous function in the long-range propagation of polarity information. Uncharacterized mechanisms of intercellular signaling then lead to autonomous activation of fz and assembly of asymmetric polarity protein complexes. Note is taken of the contrast with the recent suggestion that ds and ft act directly upstream of the autonomous function of fz (H. Strutt, 2002).

Other factors or mechanisms must also be involved in nonautonomous propagation of polarity information, in order to explain all of the observations. For instance, complete loss of fj function does not result in a loss of polarity patterning in the wing, indicating that there must be other upstream patterning factors. Furthermore, clones of fj and ft give stronger nonautononomous phenotypes in a central portion of the wing, whereas ds and fz seem to give rather similar phenotypes throughout. This suggests that there are other modulators of pathway activity that have region-specific effects (H. Strutt, 2002).

Groups of cells lacking fj function tend to round up into tight foci, appearing to have greater affinity for each other than for their fj-expressing neighbors. Furthermore, in mutant cells abutting fj-expressing neighbors, the cadherins Ds and Ft are preferentially found at the cell junctions touching fj+ cells. These observations support the notion that one role of fj in wing patterning is to alter the adhesive properties of cells and also of the cadherins Ft and Ds. It is also noteworthy that loss of ft activity results in Ds no longer being tightly localized in the apical junctional zone of cells and that, similarly, loss of ds seems to result in reduction of apical Ft localization (H. Strutt, 2002).

It is speculated that a gradient of fj activity in the wing might lead to graded Ds/Ft activity and, hence, cell adhesion. Such a gradient of cell adhesion constitutes a possible mechanism for the long-range transmission of polarity information, although direct evidence for this is lacking. It is noteworthy that fj, ft, and ds mutations also all result in truncations of the wing on the proximodistal axis, and it is possible that this phenotype is in some way due to effects on cell adhesion (H. Strutt, 2002).

Interestingly, the effect of fj clones on Ds/Ft is cell autonomous. It was suggested that, on the basis of its amino acid sequence and in vitro studies, fj encodes a secreted factor and that this property could explain its nonautonomous phenotypes. These results indicate that at least some functions of fj are cell autonomous (H. Strutt, 2002).

Frizzled function, Dishevelled and Prickle

Clones of fz show a directional nonautonomy that has been interpreted to mean that fz mediates the intercellular communication of a polarity signal. On the other hand, dishevelled is strictly cell autonomous in clones, implying that it is involved in signal reception. To investigate whether prickle participates in both reception and transmission of a polarity signal, pkpk clones in the wing were studied. Large clones in the wing express the mutant polarity typical of that region of the wing in homozygous flies. There is an occasional nonautonomous disruption of polarity in wild-type cells adjacent to the proximal or lateral margins of a pkpk clone. In these cases, a short range perturbation aligns wild-type cells with the mutant polarity pattern. There is no clear pattern to the position of such clones, but small peninsulas of pawn (pwn+) tissue surrounded by pkpk pwn tissue tend to adopt the mutant polarity pattern. Smaller clones, induced later than 72-96 hr, do not alter the polarity of adjacent pwn+ tissue. In contrast to the autonomous behavior of pk, clones of fz tricornered (trc) cause long-range domineering nonautonomy both distal (Vinson, 1987) and lateral to the clone. Wild-type hairs are oriented toward the clone as though it is acting as a polarity `sink'. Proximal cells are also directed toward the fz trc clone, but as this is the normal orientation for wing cells no polarity changes would be expected (Gubb, 1999).

The tarsal phenotype of dsh is very similar to pkpksple13 causing duplications of the T3 and T4, segments. A more extreme phenotype including complete duplications of T2 to T4 together with a well-developed ectopic T1 joint are seen in dsh; pksple, overexpression UAS:pk+ and fz mutant strains, implying that a similar polarity signaling pathway is affected. It has been suggested that pk and fz are upstream in a signaling pathway leading to dsh (Wong, 1993 and Shulman, 1998). There are several problems with a simple linear pathway, however, and the relationship between pk and fz is unclear. (1) Neither fz nor pk is clearly epistatic to the other; rather, the double mutant (pkpk; fz) phenotype is intermediate between pkpk and fz (Gubb, 1982 and Wong, 1993). (2) Whereas fz clones cause a long-range nonautonomous disruption in surrounding tissue, pkpk clones resemble dsh clones in being almost completely cell autonomous. (Both pksple1 and pkpk-sple13 clones are cell autonomous in the eye. In contrast, the pkpk triple row bristle phenotype (that results from Sple expression in the absence of Pk) is suppressed in dsh; pkpk wings and the dsh polarity pattern is modified in dsh; pksple wings (despite pksple on its own not having a wing phenotype). (3) Complete lack of pk transcripts produces a phenotype very similar to dsh. Taken together, these results indicate that Pk is not downstream of Fz but may represent an alternative input into Dsh-mediated planar polarity signaling. If Dsh is acting as a scaffolding or adapter protein then it would be required in stoichiometric ratios to its target proteins, and overexpression would prevent assembly of functional complexes. It may be that the Pk protein isoforms are components of this protein complex that are expressed in cells remodeling their cytoskeletal architecture (Gubb, 1999).

The Drosophila wing provides an appropriate model system for studying genetic programming of planar cell polarity (PCP). Each wing cell respects the proximodistal (PD) axis; i.e., it localizes an assembly of actin bundles to its distalmost vertex and produces a single prehair. This PD polarization requires the redistribution of Flamingo (Fmi), a seven-pass transmembrane cadherin, to proximal/distal cell boundaries; otherwise, the cell mislocalizes the prehair. Achievement of the biased Fmi pattern depends on two upstream components in the PCP signaling pathway: Frizzled (Fz), a receptor for a hypothetical polarity signal, and an intracellular protein, Dishevelled (Dsh). In this study, endogenous Dsh was visualized in the developing wing. A portion of Dsh colocalizes with Fmi, and the distributions of both proteins are interdependent. Furthermore, Fz controls the association of Dsh with cell boundaries: this association is correlated with the presence of hyperphosphorylated forms of Dsh. These results, together with studies on Fz distribution, support the possibility that Fz, Dsh, and Fmi constitute a signaling complex and that the restricted localization of this complex directs cytoskeletal reorganization only at the distal cell edge (Shimada, 2001).

Dishevelled was visualized in the developing wing by using specific antibodies. Some of the Dsh molecules are present at cell-to-cell boundaries in third-instar wing discs and in wings 2 hr after puparium formation (hr APF). By 18 hr APF, a larger fraction of Dsh molecules appear to be associated with cell boundaries, and then they are redistributed preferentially at proximal/distal (P/D), but not anterior/posterior (A/P), boundaries. This asymmetrical pattern is detectable at 24 hr APF and is most prominent at 30 hr APF, just prior to the onset of prehair outgrowth. This conversion produces a zigzag pattern on the epidermal plane: this pattern is highly reminiscent of the distribution of Flamingo (Fmi), a seven-pass transmembrane cadherin. In fact, Dsh and Fmi appeared to colocalize until 30 hr APF, and in terms of the time course, generation of the asymmetrical pattern of Dsh is indistinguishable from that of Fmi. Dsh, like Fmi, was present apically along the apicobasal axis at cell boundaries; curiously, Dsh distribution appears to be more restricted than that of Fmi. Besides Dsh molecules at cell boundaries, diffuse or punctate signals are also seen in the cytoplasm. Once prehairs emerge and initiate outgrowth at around 32–34 hr APF, the Fmi pattern starts to become depolarized and becomes almost nonpolar by 36 hr APF. In contrast, the asymmetrical Dsh distribution appears to persist, and patchy signals are found at distal cell vertexes and in outgrowing prehairs (Shimada, 2001).

Dsh functions in both the PCP signaling and the canonical Wnt (Wingless) pathway, and dsh1 is a missense mutation that impairs the PCP signaling. The subcellular distribution of the Dsh1 protein was monitored and it was found to remain almost entirely in the cytoplasm and is hardly detected at cell boundaries. These findings suggest that the association of Dsh with cell boundaries is a prerequisite for its role in PCP signaling. Because Dsh functions just downstream of Fz, it was expected that Fz might be necessary for the Dsh localization at cell boundaries. In fact, boundary association of Dsh is almost lost in larval wing discs or pupal wings of fz null mutants and inside fz clones. This observation is consistent with a ability of Fz, in a heterologous system, to recruit Dsh from cytoplasmic vesicles to cell boundaries (Shimada, 2001).

Fmi localization has been studied under various genetic conditions of fz that alter polarity. Dsh was examined under the same genetic conditions and it was found that Dsh, like Fmi, is redistributed to cell boundaries where there is an imbalance of Fz activity. One example was seen along borders of clones of cells homozygous for fzR52, a strong fz allele. Another example was shown in an experiment of graded fz expression. A fz gradient expression along the anterior-posterior axis of the wing reorients hairs from high to low levels of fz expression. In these wings, ectopic Dsh and Fmi accumulation at A/P cell boundaries, instead of P/D ones, prefigures prehair outgrowth in the anterior or posterior direction. A tight coupling of Dsh and Fmi mislocalization with altered polarity is also seen in areas distal to fzR52 clones. All of these results are consistent with the ideas that (1) an imbalance of Fz activity at boundaries is necessary and sufficient to localize Dsh and Fmi there, and (2) in the wild type, distributions of the two proteins at the P/D boundaries direct the cell to choose the distal edge for prehair development. Nevertheless, to rigorously demonstrate that the distributions of Dsh and Fmi play instructive roles in polarizing cells, one would need to design each molecule to mislocalize at A/P boundaries and then investigate how cells are reoriented (Shimada, 2001).

To investigate how colocalization of Dsh and Fmi at P/D boundaries is controlled, the subcellular distribution of one of the two was studied in the absence of the other. Fmi distribution is nonpolar in dsh1, and the requirement of Dsh for making the Fmi pattern asymmetric was confirmed by staining dsh1 or dshV26 clones for Fmi. Fmi is uniformly present at boundaries within these clones. Curiously, the intensity of Fmi signals seems to increase in the dsh mutant cells, and this increase implies the possibility that Dsh may be involved in the destabilization of Fmi. Inside the clone and at the clone border, Fmi remains at apical cell boundaries whether Dsh is colocalized or not. This is perhaps due to the intrinsic nature of Fmi as a transmembrane molecule. In contrast, the boundary localization of Dsh is dependent on Fmi, as demonstrated by the staining of fmi mutant clones for Dsh. Fmi is missing at cell boundaries inside the clones and at clone borders, and Dsh is hardly detected at those boundaries. At boundaries outside the clones, Dsh always coexists with Fmi. These results support a reciprocal dependence between Dsh and Fmi for subcellular localization until prehair formation and may imply a complex formation. Under the experimental immunoprecipitation conditions, however, evidence could not be provided for a physical association between Dsh and Fmi (Shimada, 2001).

Although it has been shown that Dsh is phosphorylated in response to Wingless in a cell culture system and in the embryo, posttranslational modification of Dsh has not been studied in the context of PCP signaling. Western blot analysis shows that a fraction of Dsh molecules in pupae is hyperphosphorylated and that those forms are hardly detectable in fz or fmi mutants. Thus, the absence of the hyperphosphorylated forms correlates with the loss of Dsh at cell boundaries and suggests that the hyperphosphorylation is either required for, or is a downstream readout of, the cell boundary localization. Besides reduction in the level of the hyperphosphorylated forms, the overall amount of Dsh also appears to decrease in fz or fmi mutants; an exception may be the most quickly migrating band, which most likely represents an unphosphorylated or poorly phosphorylated form (Shimada, 2001).

Where is Fz localized within the cell? It has been shown recently that the ubiquitous expression of Fz-GFP rescues a fz polarity defect and that Fz-GFP colocalizes with Fmi: these findings strongly suggest that endogenous Fz assembles with Dsh and Fmi at the P/D boundary. Furthermore, Fz-GFP distribution is regulated by Dsh and Fmi. Therefore, in the sense of subcellular localization, there seems to exist an interdependence between any two of Fz, Dsh, and Fmi. This triangular relationship can be summarized as follows: (1) in the absence of Fz or Fmi function, the intracellular protein Dsh cannot be attached to cell boundaries; (2) without Dsh function, boundary distributions of Fz and Fmi no longer become asymmetric along the P/D axis; and (3) Fz localization at cell boundaries is abolished by a loss of Fmi (Shimada, 2001).

Other data strongly suggest bilateral distribution of Fmi at the P/D boundary, and it is pointed out that this bilateral pattern per se does not explain how the distal cell vertex, not the proximal one, is selected for prehair formation. Importantly, Fz-GFP localization is unipolar; i.e., it is present only at the distal boundary. A distal concentration of Dsh molecules could be likely because Dsh positively relays Fz signaling, although the possibility of the bilateral localization still exists. It had been expected that one could answer this question by tracing Dsh signals in wild-type cells contacting proximal and distal borders of dsh mutant clones; what underlies this approach is the fact that dsh controls PCP in a cell-autonomous fashion. Given that Dsh is localized only at the distal cell edge and that dsh mutant cells do not affect Dsh localization in wild-type neighbors, Dsh signals could have been detected at interfaces between wild-type and mutant cells only along proximal borders of the clones. Unexpectedly, Dsh was not always localized at those cell boundaries along proximal borders; furthermore, 50% of wild-type cells in direct contact with proximal borders mispositioned Dsh at anterior/posterior boundaries. Therefore, dsh mutations appear to exert a one-cell nonautonomous effect on Dsh distribution, and this did not allow a conclusion about unipolar versus bilateral localization of Dsh. This one-cell nonautonomy could be due to misplaced Fz and Fmi molecules in dsh mutant cells, which might send an illegitimate message to wild-type neighbors. In any case, the local assembly of a tripartite signaling complex of Fz, Dsh, and Fmi in the cell most likely amplifies Fz signaling only at the distal cell vertex and induces cytoskeletal reorganization (Shimada, 2001).

It should be noted that in larval wing discs and early pupal wings, Dsh distribution is not asymmetric along the presumptive PD axis; nonetheless, it is associated with cell boundaries, and this association is dependent on Fz. Functional relevance of the cell boundary localization of Dsh has been suggested in vertebrate embryos, in which Dsh controls cell polarization in convergent extension movements. A fusion protein of a Xenopus Dsh homolog and GFP (Xdsh-GFP) is associated with boundaries in cells undergoing morphogenetic movement, but this protein remains in the cytoplasm of cells that are not undergoing such movement. It would be interesting to examine whether the boundary association of Xdsh-GFP or endogenous Xdsh, as well as distributions of Xenopus homologs of Fz and Fmi, are biased toward the direction of the cell movement (Shimada, 2001).

Planar cell polarity signaling in Drosophila requires the receptor Frizzled and the cytoplasmic proteins Dishevelled and Prickle. From initial, symmetric subcellular distributions in pupal wing cells, Frizzled and Dishevelled become highly enriched at the distal portion of the cell cortex. A Prickle-dependent intercellular feedback loop is described that generates asymmetric Frizzled and Dishevelled localization. In the absence of Prickle, Frizzled and Dishevelled remain symmetrically distributed. Prickle localizes to the proximal side of pupal wing cells and binds the Dishevelled DEP domain, inhibiting Dishevelled membrane localization and antagonizing Frizzled accumulation. This activity is linked to Frizzled activity on the adjacent cell surface. Prickle therefore functions in a feedback loop that amplifies differences between Frizzled levels on adjacent cell surfaces (Tree, 2002).

Using a glutathione-S-transferase (GST) pull-down assay, it was found that full-length, in vitro translated PkSple (one of the three alternatively spliced forms of Prickle) binds to a GST-Dsh fusion protein. To determine the domain dependency of this interaction, domains of Pk were tested for binding to individual Dsh domains. A Pk construct containing just the conserved PET and LIM domains that are common to all Pk isoforms (PkPETLIM), but not either domain individually, is sufficient for binding to full-length Dsh. Furthermore, of the three defined Dsh domains, DIX, PDZ, and DEP, it was found that PkPETLIM binds a construct containing the DEP domain but not other Dsh domains. The binding between Pk and Dsh was confirmed, and the domain specificity was verified, using a yeast two-hybrid assay. PkPETLIM fused to a DNA binding domain interacts specifically with the Dsh DEP domain but not with the other domains of Dsh. It is concluded that Pk, via its conserved PET and LIM domains, binds the Dsh DEP domain (Tree, 2002).

Since the Dsh DEP domain is required for membrane localization, it was hypothesized that Pk antagonizes Fz activity by interfering with Dsh membrane localization. To test this, it was necessary to find conditions in which Pk activity is uncoupled from Fz/Dsh activity in the adjacent cell. Fz-dependent Dsh localization was therefore reconstituted in cultured cells. In U20S cells, transfected GFP-tagged Dsh (Dsh::GFP) can be seen in punctate patches in the cytoplasm. On cotransfection with Fz, Dsh::GFP is translocated to the cell membrane, as was seen in a similar frog animal cap assay. To test whether the physical interaction between Pk and Dsh affects Dsh membrane localization, Fz, Dsh::GFP, and Pk were cotransfected. Dsh is localized in the cytoplasm in 90% of the cotransfected cells, resembling cells transfected with Dsh::GFP alone. In contrast, Pk lacking the PET domain is severely impaired in its ability to block Dsh membrane localization. Pk, therefore, interferes with Dsh membrane association in this heterologous system and by extension may interfere with Dsh membrane localization at the proximal boundary of pupal wing cells. Since Dsh is required to generate asymmetric localization of Fz, it is suggested that blocking Dsh membrane localization also inhibits the accumulation of Fz at the proximal boundary (Tree, 2002).

Fz activity on the distal side of the cell is linked to the accumulation of Pk on the proximal side of the adjacent cell and that accumulation of Pk suppresses Fz/Dsh localization at the proximal cell cortex. In this way, Fz and Pk appear to function in a feedback loop rather than a linear signaling pathway. The ordering of these members of the PCP pathway was analyzed through epistasis analysis. The loss-of-function phenotypes of pkpk-sple, dsh, and fz are very similar, suggesting that they activate the same signaling mechanism. However, these similarities make classical genetic epistasis tests difficult to interpret. In previous work, Fz overexpression phenotypes have been used to show that Fz signaling requires Dsh. These results were interpreted as implying that the PCP genes form a linear pathway with Dsh acting downstream of Fz. Consistent with this is the finding that Dsh localization requires Fz protein. In contrast, other observations do not fit this model. For example, localization of Fz requires Dsh, a presumed 'downstream' element of the pathway. Fz localization also requires the other presumed downstream elements Pk, the 7 cadherin Flamingo (Fmi), and novel TM protein Van gogh/Strabismis (Vang/Stbm) (Tree, 2002).

To clarify these results, several further epistasis tests were performed. Dsh was found to be required to generate a Fz overexpression phenotype and Fz is required to produce a Dsh overexpression phenotype. In addition, Pk is required for the Dsh and Fz overexpression phenotypes. These results argue against a simple linear pathway but are consistent with Fz, Dsh, and Pk participating in a feedback loop (Tree, 2002).

Evidence is provided that Fmi acts with Pk on the proximal side of the cell. Like Pk, overexpression of fmi using ptc-Gal4 causes hairs to point toward the midline of the wing. fmi overexpression clones show distal-domineering nonautonomy, phenocopying fz loss-of-function and pk overexpression clones. Thus, it is possible that Fmi could be acting similarly to Pk in feedback amplification of PCP signaling. Fmi is likely to be present at both sides of the cell, but it may act with Pk at the proximal side to suppress Fz signaling. However, Fmi must have an additional function since, within and surrounding fmi clones, very little Pk and Dsh localize to cell boundaries. In addition to functioning with Pk to suppress Fz/Dsh activity, it is suggested that Fmi could have a role in stabilizing the Fz complex at the membrane, without which no signaling occurs (Tree, 2002).

Whether differences in Pk activity between adjacent cells affect the localization of Dsh and Fz was examined. pk was overexpressed in the posterior of the wing (using UAS-pk, engrailed-Gal4), and Dsh and Fz localization were assessed near the edge of the pk overexpression domain. Dsh and Fz were relocalized to the antero-posterior cell boundaries, indicating that accumulation of Dsh and Fz occurs at interfaces between cells expressing high and low levels of Pk. Furthermore, within the posterior domain, Dsh and Fz are seen to accumulate to higher levels than in the anterior, consistent with a role for Pk in promoting PCP signaling. Two conclusions are drawn from these observations: (1) Dsh and Fz localization occur preferentially at boundaries between cells where one cell expresses high levels and the other expresses lower levels of Pk; (2) providing high levels of Pk amplifies Fz signaling, as measured by the increased accumulation of both Fz and Dsh at cell peripheries. In the wild-type, therefore, Pk at the proximal side of the cell drives Fz and Dsh accumulation at the distal side of the adjacent cell. Conversely, providing high levels of Fz induces higher levels of Pk accumulation (Tree, 2002).

Since Pk accumulates at boundaries between cells expressing different levels of Fz, and Dsh accumulates at boundaries between cells expressing different levels of Pk, it is suggested that Fz and Dsh on one side and Pk on the opposite side of an intercellular boundary form a self-organizing complex. Indeed, evidence is found for this in the posterior, pk-overexpressing domain of the same wings. In the posterior, Dsh and Fz accumulate in discrete patches around the cell peripheries. Simultaneous staining for Pk, Dsh, and Fmi, or Fz and Fmi, reveals that all four proteins colocalize to these patches. Thus, Pk, Dsh, Fmi, and Fz form self-organizing complexes bridging adjacent cells (Tree, 2002).

Various models have been proposed for the distribution of PCP information within Drosophila epithelia. Regardless of which model is correct, it is likely that this cue induces only a slight asymmetry within each cell. In the Drosophila wing, this initial small asymmetry may result in slightly higher levels of Fz accumulation or activation on the distal than on the proximal side of cells. However, this initial level of asymmetry is insufficient to polarize the cytoskeletal architecture of the cell and is undetectable as assessed by Dsh localization. Fz and Dsh localization subsequently become highly polarized through the action of a feedback amplification loop. Although the mechanism for such a feedback loop has not been demonstrated, it has previously been speculated that Fz activity on distal cell surfaces might promote expression of an inhibitory Wnt molecule that would suppress Fz activity on the adjacent cell surface. This study demonstrates suppression of Fz activity on adjacent cell surfaces but finds that it is mediated through asymmetric localization of Pk, which antagonizes Fz/Dsh accumulation by blocking cortical Dsh localization (Tree, 2002).

A model is proposed for a Pk-dependent feedback mechanism that functions across each P-D cell interface to amplify the difference between the initial levels of Fz/Dsh activity. Pk and Dsh are initially distributed nearly symmetrically, but in largely nonoverlapping distributions around the cells' apices. As a result of an initial, slight asymmetry in one or more components, Pk suppresses Fz accumulation on the proximal side of the cell by antagonizing Dsh cortical localization. Dsh association with the cortex is required for asymmetric Fz accumulation. Reduced proximal Fz accumulation then decreases Pk activity on the adjacent, distal side of the neighboring cell, allowing even greater Dsh cortical localization and Fz accumulation. Conversely, Pk localization is induced by Fz localization on the neighboring cell. The result is the highly asymmetric distribution of Fz, Dsh, and Pk observed at 30 hr APF. This mechanism is bidirectional, constituting a negative feedback loop that is predicted to be unstable; once any asymmetry is initiated, high levels of Fz/Dsh are promoted on one side of the boundary and low levels on the other. Fmi is proposed to play two roles. The localization of all components to the apical cell cortex depends on Fmi, which is thought to be on both sides of the boundary. Since Fmi overexpression mimics Pk overexpression, it is proposed that Fmi is activated selectively on the proximal side of the cell, where it works with Pk to suppress Fz/Dsh activity. In essence, the function of this regulatory loop depends on the balance of forces regulating Dsh membrane association on either side of the boundary. Fz recruits Dsh to the cell cortex, and Pk blocks this recruitment. Once an initial asymmetry is induced, these forces become different on either side of the boundary, and the feedback mechanism amplifies the differences (Tree, 2002).

Because fmi overexpression produces polarity phenotypes very similar to those seen with pk overexpression, it is likely that Fmi activity is also polarized across cell boundaries and participates in blocking Fz/Dsh accumulation on the distal side of the interface. However, Fmi has been proposed to exist on both sides of the boundary, and Fmi is required for the efficient localization of Fz, Dsh, and Pk to boundaries. Fmi, therefore, appears to have an additional, complex-stabilizing function. Diego (Dgo), an ankyrin repeat protein, is required for PCP signaling and localizes to P-D cell boundaries, though it has not been possible to resolve on which side of the boundary Diego is found. Dgo may function as a scaffold for assembly of PCP-signaling components (Tree, 2002).

The core PCP protein Van Gogh (Vang, also known as Strabismus [Stbm], a transmembrane protein, is likely to be involved in the feedback amplification mechanism. In a vang/stbm mutant background, Fz is localized symmetrically to the membrane in the same manner as in a pk null background, and vang/stbm has been proposed to function downstream of pk. Interestingly, in Xenopus and zebrafish, Stbm binds Dsh and seems to toggle its activity between the canonical Wnt and a PCP-like signaling pathway. Stbm antagonizes canonical Wnt signaling and activates a PCP-like pathway that regulates convergent extension and induces JNK signaling. It is proposed that both fly and vertebrate Vang/Stbm may function together with Pk, facilitating PCP signaling by antagonizing Dsh activity (Tree, 2002).

The feedback amplification mechanism described here provides a clue to understanding the long-standing problem of domineering nonautonomy. Loss-of-function clones of fz, vang/stbm, and to a lesser extent pk induce polarity phenotypes in neighboring wild-type tissue. The feedback loop model predicts that loss of Fz will disrupt the localization of these components in the neighboring distal cells, causing their polarity to reverse. Furthermore, if the immediate clone neighbors have reversed polarity, this could then cause the reversal to propagate over a distance. This phenomenon is observed near the lateral borders of a fz clone, where Pk accumulation is seen to run parallel to the clone, even at a distance from the mutant cells. Thus, the reversal in Pk localization may underlie, in part, the domineering nonauontomy observed within wild-type tissue distal to fz clones (Tree, 2002).

Once asymmetrically localized, the Fz/Dsh complex must direct reorganization of the cytoskeleton by both determining the location for prehair initiation and by limiting the number of prehairs to one. Rho and Rho-associated kinase directly regulate myosins to limit the number of prehairs. The recently described Formin homology protein, Daam1, links Dsh to Rho during vertebrate PCP-like signaling, and a Drosophila homolog may function similarly. However, further studies will be required to understand how localized Fz and Dsh orient prehair initiation (Tree, 2002).

The Drosophila epidermis is characterized by a dramatic planar/tissue polarity. The frizzled pathway has been shown to be a key regulator of planar polarity for hairs on the wing, ommatidia in the eye, and sensory bristles on the notum. The genetic relationships between putative frizzled pathway downstream genes inturned, fuzzy, and multiple wing hairs (inturned-like genes) and upstream genes such as frizzled, prickle, and starry night (frizzled-like genes) were investigated. Previous data have shown that the inturned-like genes are epistatic to (function downstream of) the frizzled-like genes when the entire wing is mutant. Consistent with this are observations showing that the asymmetric accumulation of Fz, Dsh, and Fmi is not altered in in, fy, or mwh mutants. Those experiments were extended and the behavior of frizzled clones in mutant wings was examined. The domineering nonautonomy of frizzled clones is not altered when the clone cells are simultaneously mutant for inturned, multiple wing hairs, or dishevelled but it is blocked when the entire wing is mutant for inturned, fuzzy, multiple wing hairs, or dishevelled. Thus, for the domineering nonautonomy phenotype of frizzled, both inturned and multiple wing hairs are needed in the responding cells but not in the clone itself. Expressing a number of frizzled pathway genes in a gradient across part of the wing repolarizes wing cells in that region. inturned, fuzzy, and multiple wing hairs are required for a gradient of frizzled, starry night, prickle, or spiny-legs expression to repolarize wing cells. These data argue that inturned, fuzzy, and multiple wing hairs are downstream components of the frizzled pathway. To further probe the relationship between the frizzled-like and inturned-like genes the consequences of altering the activity of frizzled-like genes was determined in wings that carried weak alleles of inturned or fuzzy. Interestingly, both increasing and decreasing the activity of frizzled and other upstream genes enhances the phenotypes of hypomorphic inturned and fuzzy mutants. Results from several different experimental paradigms support the hypotheses that in and fy function downstream of fz, dsh, stan, pk, and sple in the wing and that they are required for the transduction of the fz signal to regulate hair morphogenesis and hence the actin and microtubule cytoskeletons (Lee, 2002).

To determine the relative positions of in, fy, and mwh with respect to several other fz-like tissue polarity genes, their ability to block the gain-of-function phenotypes that result from the directed expression of stan/fmi, pk, and sple were examined. ptc-GAL4 was used to overexpress pk. This results in a band of expression in the C region of the wing, with a decreasing gradient of expression away from the center of the ptc expression domain. This results in hairs pointing toward the midline in the C region of the wing. This phenotype is blocked by mutations in in, fy, and mwh. The resulting wings (e.g., ptc-GAL4/UAS-pk; in/in) show no effects from the directed expression of pk. Two different GAL4 drivers (ms1096-GAL4 and act-GAL4) were used to direct the expression of UAS-sple. Both of these drivers result in relatively even expression across the wing (although in ms1096 the level in the dorsal cell layer is higher than that in the ventral cell layer). This results in a reversal of hair polarity over much of the wing that resembles that seen in pkD. Once again it was found that mutations in in, fy, and mwh are able to block this gain-of-function phenotype. The omb-GAL4 driver was used to drive expression of UAS-stan. In the wing disc omb-GAL4 drives expression in a band located centrally along the anterior/posterior axis of the wing. However, in the pupal wing the expression pattern is more complicated during the time for tissue polarity development. During this time in the distal region of the wing, a series of alternating bands of expression and no expression is seen. Driving expression of stan using omb-GAL4 leads to a series of bands of polarity reversals; polarity is oriented from cells of low toward cells of high stan expression. Mutations in in, fy, and mwh are able to block the consequences of overexpressing stan. These data argue that in, fy, and mwh functioned downstream of fz, pk, sple, and stan in tissue polarity (Lee, 2002).

Driving expression of dsh using the ptc-GAL4 driver results in the formation of multiple hair cells in the proximal part of the ptc expression domain. In addition, ectopic bristles are also often found, presumably due to the role of dsh in canonical wg signaling. When UAS-dsh +/+ ptc-GAL4; in/in flies were examined, an additive multiple hair cell phenotype was found (i.e., stronger than either single phenotype) in the proximal part of the ptc domain and the presence of ectopic bristles. Equivalent results were found in UAS-dsh fy; fy ptc-GAL4 wings. The formation of the ectopic bristles was not surprising, since this is likely an effect of dsh on canonical wg signaling and in and fy are not thought to play a role in wg signaling. The additive multiple hair cell phenotype was unexpected and argues either that in and fy are not downstream of dsh in the fz pathway or that when overexpressed, Dsh can bypass the requirement for in and fy function (Lee, 2002).

The overexpression of fz just prior to hair initiation results in a multiple hair cell phenotype, which resembles that of in and fy (this is sometimes referred to as the late fz gain of function). When fz was overexpressed just prior to hair initiation in pupae that were mutant for a null allele of either in or fy an increased number of multiple hair cells was found. Thus, for the late fz gain of function, as for the overexpression of dsh, either in and fy are not required for fz signal transduction or when fz is overexpressed late the requirement for in and fy can be bypassed (Lee, 2002).

To test whether the frizzled-like genes could be regulating the inturned-like genes negatively or positively, the effects of both decreased and increased fz-like gene activity on hypomorphic alleles of in and fy were examined. In these experiments advantage was taken of the much higher number of multiple hair cells found in in and fy mutants compared to fz-like mutants. In all cases examined, mutations (null or hypomorphic) in fz, pk, Vang, stan, and dsh acted as strong enhancers of a weak in or fy phenotype as assayed by the frequency of multiple hair cells. These data are consistent with the fz-like genes acting as positive regulators of in and fy (Lee, 2002).

If there were a simple positive relationship between the fz-like genes and in and fy then it would be expected that the overexpression of fz-like genes should suppress hypomorphic alleles of in and fy. To test this fz was overexpressed from a hs-fz transgene and it was found that this enhances the multiple hair cell phenotypes of weak inII53 and fyJN12 alleles. These results argued that fz antagonizes the activity of in and fy. The genetic relationship between pk, sple, and fmi and weak alleles of in and fy was similarly tested. Overexpression of sple and stan from ms1096-GAL4 (or act-GAL4); UAS-sple and omb-GAL4; UAS-stan, respectively, in hypomorphic mutant backgrounds of in and fy results in an increase in the multiple hair cell phenotypes in the region of the wing, where expression is driven by the GAL4 enhancer trap. A similar enhancement was not seen for the overexpression of pk from ptc-GAL4 UAS-pk. These results suggest that sple and stan antagonize the activity of the in and fy. Notably, the gain-of-function polarity phenotypes of overexpressed pk, sple, and stan are blocked even with weak alleles of in and fy, confirming that these genes are required for the function of the fz pathway. It is clear that the fz-like genes do not act as simple positive or negative regulators of in and fy. The interaction is reminiscent of the observations that similar phenotypes result from either overexpression or a lack of function of fz-like genes (Lee, 2002).

The disruption of the microtubule cytoskeleton by treatment with vinblastine or colchicine results in many wing cells forming more than one hair. At the level of the individual cell these multiple hair cells resemble those found in in and fy mutants since the hairs are formed at the cell periphery and appear to be the result of independent initiation events. The altered polarity seen in in and fy mutant cells is not induced by vinblastine or colchicine. Immunostaining has not revealed any defects in microtubule organization in in or fy mutants. Thus, it seems unlikely that the in and fy phenotypes are due to disrupting the microtubule cytoskeleton and in this way indirectly affecting hair morphogenesis. However, it seemed possible that the disruption of the microtubule cytoskeleton could be inducing multiple hairs by interfering with in and/or fy function. It was reasoned that if that is the case then treatment of in or fy null mutant wing cells with vinblastine or colchicine would not produce a stronger phenotype than the in or fy null cells by themselves. Vinblastine was injected into in or fy mutant pupae and an additive response was found. That is, injected in or fy mutant cells produces a stronger multiple hair cell phenotype than does untreated mutant cells or cells from injected wild-type pupae. This result shows that the microtubule disruption phenotype cannot be due to a simple interference with in/fy function and argues that the microtubule cytoskeleton has a function that is independent of the fz pathway. The additive response in this experiment stands in sharp contrast with the lack of additivity in double mutants of in, fy, frtz, and mwh (Lee, 2002).

The hypothesis that in, fy, and mwh function downstream of fz and dsh originally came from the observation that mutations in in, fy, and mwh are epistatic to mutations in fz and dsh. Consistent with this hypothesis, the distal accumulation of Fz, Dsh, Fmi, and Dgo is not altered in in, fy, or mwh mutants. in, fy, and mwh are required for cells to either detect or respond to the lack of fz activity in a clone of neighboring mutant cells. The function of in and fy was required for cells to respond to a directed gradient of fz, stan, pk, or sple expression. Thus, using both of these experimental approaches it was found that in and fy are required for the function of the fz pathway. The simplest interpretation of this is that in and fy are essential downstream components of the fz pathway in the wing. An alternative explanation is that in and fy function in a parallel pathway and that the function of the fz pathway is dependent on this putative in pathway. There were a couple of surprising exceptional results. The overexpression of dsh or the late overexpression of fz induces the formation of multiple hair cells and it was found that these phenotypes are additive with an in or fy mutation. This could indicate that there are multiple pathways downstream of fz and dsh in the wing and that one of them does not require in or fy function. It is thought more likely that this is an example of overexpression bypassing a normal requirement for downstream proteins. It is suggested that In and Fy normally function as adapters to allow distally accumulated Fz and Dsh to stimulate the cytoskeleton and induce hair initiation in the correct part of the cell. When Fz or Dsh is overexpressed just prior to hair initiation the high cellular concentration might allow these proteins to bypass the need for In and Fy and to interact with and hyperactivate the hair initiation machinery (Lee, 2002).

In vertebrate embryos, a fz-based planar polarity pathway regulates convergent extension. Thus far there has been no evidence for a fz-based planar pathway having an analogous role in Drosophila. The eversion defect seen in fz mutant wings may be such an example. Some fz flies have a deformed wing that is short, fat, and often kinked proximally. This is seen in all strong and null fz alleles examined. This is due to a defect in wing disc eversion. In a normal pupa the wing everts so that it extends posteriorly and ventrally. In a fz mutant many pupae contain wings that extend anteriorly and sometimes dorsally. The wing eversion phenotype is another phenotype where in and fy appear to be epistatic to fz. The observations are somewhat puzzling as the defect in fz appears to be rescued by a mutation in in or fy. Thus, it is not clear that a simple inactivation of the fz pathway is responsible for the eversion phenotype. Perhaps for this phenotype there are multiple inputs into the fz pathway upstream of in and fy and the eversion defect is due to an imbalance in the input to in and fy. Blocking the pathway at in or fy would eliminate the function of the pathway completely and suppress miseversion (Lee, 2002).

The cellular mechanisms involved in the eversion of the wing are only poorly understood, but it is possible that convergent extension plays a role. It is also possible that the joint defects seen in fz mutants have a similar basis. In vivo observations on everting wings and joint morphogenesis would make an important contribution toward determining if convergent extension plays a role in these morphogenetic events. The development of in vivo imaging approaches for Drosophila pupae makes such experiments possible (Lee, 2002).

CKIε/discs overgrown promotes both Wnt-Fz/β-catenin and Fz/PCP signaling in Drosophila

The related Wnt-Frizzled(Fz)/β-catenin and Fz/planar cell polarity (PCP) pathways are essential for the regulation of numerous developmental processes and are deregulated in many human diseases. Both pathways require members of the Dishevelled (Dsh or Dvl) family of cytoplasmic factors for signal transduction downstream of the Fz receptors. Dsh family members have been studied extensively, but their activation and regulation remains largely unknown. In particular, very little is known about how Dsh differentially signals to the two pathways. Recent work in cell culture has suggested that phosphorylation of Dsh by Casein Kinase I ε may act as a molecular 'switch', promoting Wnt/β-catenin while inhibiting Fz/PCP signaling (Cong, 2004). This study demonstrates in vivo in Drosophila through a series of loss-of-function and coexpression assays that CKIε acts positively for signaling in both pathways, rather than as a switch. The data suggest that the kinase activity of CKIε is required for peak levels of Wnt/β-catenin signaling. In contrast, CKIε is a mandatory signaling factor in the Fz/PCP pathway, possibly through a kinase-independent mechanism. Furthermore, the primary kinase target residue of CKIε on Dsh has been identified. Thus, the data suggest that CKIε modulates Wnt/β-catenin and Fz/PCP signaling pathways via kinase-dependent and -independent mechanisms (Klein, 2006).

Cell-culture assays have suggested that CKIε positively regulates Wnt-Fz/β-catenin signaling and that it antagonizes Fz/PCP signaling. To confirm that CKIε is required for Wnt/β-catenin signaling in vivo, loss of function (LOF) alleles of discs overgrown (dco/doubletime, the Drosophila CKIε gene) were examined for phenotypes indicative of Wingless signaling defects. Consistent with previous data demonstrating a requirement for dco in disc growth, strong dco/CKIε alleles (dcodbt-P) gave clones too small to analyze for disruption of Wg target-gene expression. Thus dco/CKIε mutant clones were generated with the Minute technique. In these, expression of the Wg target gene senseless (sens) was lost in mutant cells. Accordingly, dco/CKIε adult wing clones show loss of margin bristles and/or parts of the wing margin, consistent with a positive requirement for dco in Wg signaling. However, Wg targets that require lower levels of Wg signaling (e.g., Dll) were not affected, indicating that dco/CKIε is only required for peak Wg signaling levels. Consistent with this finding, genome-wide RNAi screens for Wg signaling components identified dco/CKIε as a factor required for peak levels of β-catenin reporter expression (Klein, 2006).

The Fz/PCP pathway can easily be studied in Drosophila. The precise ommatidial arrangement in the eye and the orientation of hairs on the wing depend on correct Fz/PCP input. The two best-studied PCP signaling factors are Fz and Dsh, which also act in canonical Wnt/β-catenin signaling. To study the role of dco/CKIε in PCP, various heteroallelic dco combinations were analyzed. In several of these (e.g., dcodbt-P/dcodbt-AR), typical PCP defects were seen. Clones of a strong dco/CKIε allele show classical PCP phenotypes in the wing, with reoriented wing hairs, and in the eye, with ommatidal chirality and orientation defects (dco/CKIε clones also displayed ommatidia with photoreceptor loss, likely as a result of the cell viability requirement of Dco/CKIε (Klein, 2006).

In vivo LOF analyses allow led to the conclusion that dco/CKIε is required for peak levels of Wg signaling, but does not appear to be a mandatory Wnt/β-catenin signaling component. In addition, the data identify dco/CKIε as a new factor required in Fz/PCP signaling (Klein, 2006).

To dissect the function of dco/CKIε in Wg and PCP signaling, the effects of overexpressing CKIε were examined. In the eye, dco/CKIε was overexpressed with sevenless(sev)-Gal4 (in R3/R4 cells that are critical for PCP establishment, which causes PCP phenotypes. In the wing, decapentaplegic(dpp)-Gal4-driven expression in a proximal-distal stripe along the A-P compartment boundary can be used to identify positive and negative effects on both Wg and PCP signaling. dpp>CKIε displayed a mild but consistent PCP defect, a characteristic hair swirl near the intersection of the dpp stripe and wing margin. It also caused a small number of extra margin bristles, typical of increased Wg signaling (Klein, 2006).

Given that these phenotypes were mild, attempts were made to enhance them. Because CKIε requires an activating dephosphorylation event, which can be induced by Fz signaling, the effect was tested of coexpressing CKIε with either Fz or Fz2 in the dpp stripe. Overexpression of Fz (dpp>Fz) causes a characteristic reorientation of wing hairs that point away from the expression domain, but does not induce ectopic margin bristles. Strikingly, coexpression of Fz and CKIε leads to a dramatic synergy, with enhanced PCP defects and a large number of extra margin bristles in the expression domain, indicative of a positive CKIε role in both Wg and PCP signaling. Consistently, expression of Fz or CKIε alone is not sufficient to induce visible changes in Wg target-gene expression, but coexpression of CKIε and Fz cell-autonomously induces ectopic Senseless-positive cells (Klein, 2006).

dpp>Fz2 induces many ectopic margin bristles near the intersection of dpp expression and the wing margin (at 25°C;). Because dpp-driven coexpression of Fz2 and CKIε at 25°C is lethal, the dpp>Fz2, CKIε coexpression was examined in flies raised at 18°C (allowing for weaker Gal4-driven expression). dpp>Fz2 at 18°C induces only few margin bristles, a phenotype that is enhanced upon coexpression with CKIε; dpp>Fz2, CKIε coexpression also induces PCP-like hair swirls, although this effect could be indirect given that this combination induces wing-margin-like vein tissue, which could repolarize parts of the wing blade). Taken together with the LOF analyses, these data demonstrate a positive, synergistic role for Dco/CKIε not only in Wg signaling, but also in Fz/PCP signaling (Klein, 2006).

To confirm that dco/CKIε acts positively for both pathways in vivo, genetic interactions of dco with known Wg and PCP signaling factors were examined in the eye. Overexpression of Fz (sevFz) causes strong PCP defects, a phenotype that is significantly suppressed by the removal of a single copy of dco. Overexpression of Strabismus (Stbm; also known as Van Gogh or Vang), an antagonist of Fz/PCP signaling, with sevGal4 (sevStbm) causes mild PCP defects, which are enhanced by removal of a copy of dco, again supporting a positive role for dco in Fz/PCP signaling (Klein, 2006).

The effect of removing a copy of dco in the context of Dsh overexpression (sev>Dsh). sev>Dsh causes both PCP defects and loss of photoreceptors, with the latter resembling the effect of sev>Wg. Removing a copy of dco strongly suppresses the loss-of-photoreceptor phenotype, supporting a positive role for dco for peak Wg signaling. Assessing the effect of the removal of a copy of dco on the PCP in sev>Dsh is not possible, because the loss-of-photoreceptor phenotype masks PCP defects in many ommatidia (Klein, 2006).

In addition, genetic analyses was performed with overexpressed CKIε. sev>CKIε eyes exhibit mild PCP defects and a small percentage of ommatidia with a change in photoreceptor number. Removal of a single copy of dsh suppressed the PCP and photoreceptor number defects, consistent with Dco/CKIε acting positively together with Dsh to elicit both phenotypes. Consistent with the coexpression results in the wing, these interactions support a positive role for dco/CKIε in both Wg and Fz/PCP signaling (Klein, 2006).

On the basis of the genetic interaction data with Dsh and cell-culture and in vitro kinase assays that have shown that CKIε can bind and phosphorylate Dsh, Dsh appears to be a likely phosphorylation target of CKIε. The actual site of phosphorylation on Dsh, however, has not been mapped. To narrow down the region of phosphorylation, a series of GST-Dsh constructs were generated covering all domains. In vitro kinase assays using CKIε showed specific phosphorylation of all Dsh isoforms containing the basic region and PDZ domain (GST-Dsh, GST-bPDZ, GST-ΔC), but not of those containing just the PDZ (GST-PDZ) or other parts of Dsh (Klein, 2006).

To determine the exact site of phosphorylation, the region N terminal to the PDZ domain was analzyed for conserved CKIε consensus sites, and a likely motif was identified. In this motif, S236 is predicted to be the first serine residue phosphorylated by CKIε. When this serine is mutated to alanine (within GST-bPDZ), CKIε is no longer able to phosphorylate Dsh. Interestingly, this residue is in a short region of Dsh that is important for Dsh phosphorylation, activity, and signal specificity (Klein, 2006).

Because previous studies have suggested that CKIε kinase activity is required for its ability to transduce Wnt/β-catenin signals, a kinase-dead isoform of CKIε, Dco/CKIεD132N (which affects the ATP binding site) was tested in vivo. This mutant was unable to transduce Wg signals, but, surprisingly, it induced strong GOF PCP phenotypes in both the eye and wing. sev>CKIεD132N eyes display clean PCP phenotypes. Wings from dpp>CKIεD132N flies also display PCP defects, with wing hairs that point away from the expression domain, demonstrating a GOF Fz/PCP phenotype (Klein, 2006).

To investigate the potential for a kinase-independent role of Dco/CKIε, Fz and CKIεD132N (with dppGAL4) were coexpressed. In contrast to coexpression of Fz and wild-type CKIε, CKIεD132N did not display extra margin bristles, indicating that kinase activity is important for Wg signaling. Similarly, in contrast to coexpression of Fz2 and wild-type CKIε, dpp>Fz2, CKIεD132N (at 18°C) did not cause an increase in margin bristles as compared to dpp>Fz2 alone. At 25°C, dpp>Fz2, CKIεD132N was not lethal, as compared to dpp>Fz2 together wild-type CKIε (Klein, 2006).

In summary, these data demonstrate a requirement for the Dco/CKIε kinase activity in Wg signaling, with CKIεD132N often acting as a dominant negative, but suggest that kinase activity is not required for Dco/CKIε activity during Fz/PCP signaling (Klein, 2006).

The CKIε requirement for Dsh phosphorylation was examined. In an unbiased Drosophila S2-cell-based screen for kinases that are required for the PCP-signaling-associated Dsh phosphorylation, dco/CKIε was identified as a kinase required in this context. Strikingly, the kinase-dead CKIε isoform, CKIεD132N, promotes PCP-signaling-associated Dsh phosphorylation as much as wild-type CKIε. These data suggest that although the presence of CKIε protein is important for this phosphorylation event, its kinase activity is not required, consistent with the in vivo expression data with Dco/CKIεD132N (Klein, 2006).

A possible caveat to these data is that, rather than acting on downstream targets in a kinase-independent manner, CKIεD132N could act to titrate away factors that inhibit the endogenous CKIε. To test this, the effect of CKIεD132N on the phosphorylation of Dsh was examined in S2 cells in the absence of endogenous CKIε. These data show that endogenous CKIε is not mediating the CKIεD132N effect, and thus CKIε is likely to act in a kinase-independent manner in the PCP context (Klein, 2006).

The in vivo data support a positive requirement of dco/CKIε for peak levels of Wnt/β-catenin signaling and a strict requirement in the Fz/PCP pathway. Whereas the kinase activity of CKIε is required for Wnt/β-catenin signaling, the analysis suggests that it is not required for Fz/PCP signaling. These findings differ from the proposed inhibitory effect of CKIε on Fz/PCP signaling in cell culture. It is possible that the PCP readout in cell culture, namely activation of JNK, reflects only a subset of PCP activities of Dsh, not representing an accurate measure of overall PCP activity. Alternatively, CKIε could act as a constitutively active kinase when expressed in cell culture, whereas its activity is regulated in vivo. Thus, for Fz/PCP signaling in vivo, the primary role for CKIε may not be as an active kinase, but rather as a stabilizer of a complex that allows for PCP-specific Dsh phosphorylation. This is supported by the data that the kinase activity of CKIε is not required for Fz/PCP signaling and that kinase-dead CKIε still stimulates phosphorylation of Dsh (Klein, 2006).

CKIε phosphorylates a specific residue in Dsh, S236, in a short region known to be phosphorylated by multiple kinases and suggested to be important in the regulation of Dsh signal specificity. This supports the proposed possibility of in vivo competitive phosphorylation as a mechanism for Dsh regulation. The region upstream of the PDZ domain appears to act as a docking site for Dsh binding proteins. Differential phosphorylation of this region could alter the binding properties of Dsh. In support of this possibility, protein-protein interaction studies have identified a large number of proteins that bind to the Dsh PDZ domain. It is unlikely that all these interactions occur at the same time, and phosphorylation is a potential mechanism to regulate this. Further experiments are needed to finely map the many potential phosphorylation target residues and the corresponding kinases and demonstrate their in vivo significance (Klein, 2006).

A signaling network for patterning of neuronal connectivity in the Drosophila brain

The precise number and pattern of axonal connections generated during brain development regulates animal behavior. Therefore, understanding how developmental signals interact to regulate axonal extension and retraction to achieve precise neuronal connectivity is a fundamental goal of neurobiology. This question was investigated in the developing adult brain of Drosophila. Extension and retraction is regulated by crosstalk between Wnt, fibroblast growth factor (FGF) receptor, and Jun N-terminal kinase (JNK) signaling, but independent of neuronal activity. The Rac1 GTPase integrates a Wnt-Frizzled-Disheveled axon-stabilizing signal and a Branchless (FGF)-Breathless (FGF receptor) axon-retracting signal to modulate JNK activity. JNK activity is necessary and sufficient for axon extension, whereas the antagonistic Wnt and FGF signals act to balance the extension and retraction required for the generation of the precise wiring pattern (Srahna, 2006).

Based on the observation that blocking Fz2 results in decreased numbers of dorsal cluster neuron (DCN) axons in the medulla, it was reasoned that Fz2 could be a receptor for a putative stabilization signal. Since Fz2 and Fz are partially redundant receptors for the canonical Wnt signaling pathway, expression of the canonical Wnt ligand Wingless (Wg) was investigated in the brain during pupation. However, no Wg expression was detected in the pupal optic lobes, suggesting that Wg is unlikely to be involved in regulating DCN axon extension. Therefore, the expression of Wnt5, which has been shown to be involved in axon repulsion and fasciculation in the embryonic CNS, was investigated. Anti-Wnt5 staining revealed widely distributed Wnt5 expression domains beginning at PF and lasting throughout pupal development and into adult life. Wnt5 is strongly expressed in the distal medulla and is also present on axonal bundles crossing the second optic chiasm.The number of DCN axons crossing to the medulla was examined in wnt5 mutant flies. The number of DCN axons crossing the optic chiasm is reduced from 11.7 to 7.9 in the absence of wnt5, suggesting that it may play a role in stabilizing DCN axons (Srahna, 2006).

Next, the requirement of the Wnt signaling adaptor protein Dsh was tested. In animals heterozygous for dsh6, a null allele of dsh, the average number of DCN axons crossing between the lobula and the medulla is reduced from 11.7 to 7.6 with 78.5% showing less than eight axons crossing. Signaling through Dsh is mediated by one of two domains. Signaling via the DIX (Disheveled and Axin) domain is thought to result in the activation of Armadillo/β-Catenin. DEP (Disheveled, Egl-10, Pleckstrin) domain-dependent signaling results in activation of the JNK signaling pathway by regulation of Rho family GTPase proteins during, for example, convergent extension movements in vertebrates. To uncover which of these two pathways is required for DCN axon extension the dsh1 mutant, deficient only in the activity of the DEP domain, was tested. Indeed, in brains from dsh1 heterozygous animals the number of extending axons was reduced from 11.7 to 7.4. In flies homozygous for the dsh1 allele the average number of axons crossing was further reduced to 4.7, with all the samples having less than six axons crossing. In contrast, the DCN-specific expression of Axin, a physiological inhibitor of the Wnt canonical pathway, did not affect the extension of DCN axons. Similarly, expression of a constitutively active form of the fly β-Catenin Armadillo also had no apparent effect on DCN extension. Finally, whether Wnt5 and Dsh interact synergistically was tested. To this end, wnt5, dsh1 trans-heterozygous animals were generated. These flies show the same phenotype as flies homozygous for dsh1, suggesting that Wnt5 signals through the Dsh DEP domain (Srahna, 2006).

To determine if dsh is expressed at times and places suggested by its genetic requirement in DCN axon outgrowth, the distribution of Dsh protein during brain development was examined. Dsh protein is ubiquitously expressed during brain development. High expression of Dsh is detected in the distal ends of DCN axons at about 15% PF shortly before they extend across the optic chiasm toward the medulla. In general, higher levels of Dsh were observed in the neuropil than in cell bodies (Srahna, 2006).

In summary, these data indicate that the stabilization of DCN axons is dependent on the Dsh protein acting non-canonically via its DEP domain. Importantly, the axons that do cross in dsh mutant brains do so along the correct paths. This suggests that, like JNK signaling, Wnt signaling regulates extension, but not guidance, of the DCN axons (Srahna, 2006).

Wnt signaling to Dsh requires the Fz receptors. To examine if the effect of Wnt5 on DCN axon extension is also mediated by Fz receptors, the number of DCN axons crossing the optic chiasm in was counted fz, fz2, and fz3 mutants. There was no significant change in the number of axons crossing in the brain of fz3 homozygous animals. In contrast, in brains heterozygous for fz and fz2, the number of the axons crossing was reduced from 11.7 to 6.6 (fz) and 6.9 (fz2), with 71% and 85.7%, respectively, showing less than eight axons crossing. These data suggest that DCN axons respond to Wnt5 using the Fz and Fz2 receptors, but not Fz3. To determine whether the Fz receptors act cell-autonomously in individual DCNs, single-cell clones doubly mutant for fz and fz2 were generated and the number of DCN axons crossing the optic chiasm was counted. In contrast to wild-type cells, where 37% of all DCN axons cross, none of the fz, fz2 mutant axons reach the medulla. To test whether wnt5, fz, and fz2 genetically interact in DCNs, flies trans-heterozygous for wnt5 and both receptors were examined. Flies heterozygous for both wnt5 and fz mutations show a strong synergistic loss of DCN axons (11.7 to 3.7) and in fact have a phenotype very similar to that of flies homozygous for dsh1. Flies doubly heterozygous for wnt5 and fz2 also show a significant decrease in DCN axons (5.7), compared with either wnt5 (~8) or fz2 (8.5) mutants. These data indicate that the genetic interaction between wnt5 and fz is stronger than the interaction between wnt5 and fz2 (Srahna, 2006).

Examination of the expression domains of Fz and Fz2 in the developing brain supports the possibility that they play roles in stabilizing DCN axons. Both Fz and Fz2 are widely expressed in the developing adult brain neuropil. In addition, Fz is expressed at higher levels in DCN cell bodies (Srahna, 2006).

The observation that the wnt5 null phenotype can be enhanced by reduction of Fz, Fz2, or Dsh suggests that another Wnt may be partially compensating for the loss of Wnt5. To test this possibility, flies heterozygous for either wnt2 or wnt4 were examined. wnt2 heterozygotes display a reduction of DCN axon crossing from 11.7 to 7.3, whereas no phenotype was observed for wnt4. Thus, wnt2 and wnt5 may act together to stabilize the subset of DCN axons that do not retract during development. In summary, these results support the model that Wnt signaling via the Fz receptors transmits a non-canonical signal through Dsh resulting in the stabilization of a subset of DCN axons (Srahna, 2006).

Data is provided that supports the hypothesis that the regulation of JNK by Rac1 modulates DCN axon extension. As such attempts were made to determine how Wnt signaling might interact with Rac1 and JNK. The opposite phenotypes of dsh and Rac1 loss-of-function suggest that they might act antagonistically. To determine if Rac1 is acting upstream of, downstream of, or in parallel to Dsh in DCN axon extension, dominant-negative Rac1 was expressed in dsh1 mutant flies. If Rac1 acts upstream of Dsh, the dsh1 phenotype (i.e., decreased numbers of axons crossing the optic chiasm) is expected. If Rac1 acts downstream of Dsh, the Rac1 mutant phenotype (i.e., increased number of axons crossing) would be expected If they act in parallel, an intermediate, relatively normal phenotype is expected. Increased numbers of axon crossing were observed, suggesting that Rac1 acts downstream of Dsh during DCN axon extension and that Dsh may repress Rac1 (Srahna, 2006).

Next, whether Dsh control of DCN axon extension is mediated by the JNK signaling pathway acting downstream of Wnt signaling was tested, as the similarity of their phenotypes suggests. If this were the case, activating JNK signaling should suppress the reduction in Dsh levels. Conversely, reducing both should show a synergistic effect. Therefore the JNKK hep was expressed in dsh1 heterozygous flies and it was found that the hep gain-of-function is epistatic to dsh loss-of-function. Furthermore, reducing JNK activity by one copy of BSK-DN in dsh1 mutant animals results in a synergistic reduction of extension to an average of 0.8 axons with 60% showing no axons crossing and no samples with more than three axons. In summary, the results of genetic analyses suggest that Wnt signaling via Dsh enhances JNK activity through the suppression of Rac1 (Srahna, 2006).

Dsh appears to promote JNK signaling and to be expressed in DCN axons prior to their extension toward the medulla early in pupal development. Since JNK signaling is required for this initial extension, it may be that Dsh also plays a role in the early extension of DCN axons. To test this possibility, DCN axon extension was examined at 30% pupal development in dsh1 mutant brains. In wild-type pupae, essentially all (~40) DCN axons extend toward the medulla. In contrast, in dsh1 mutant pupae, a strong reduction in the number of DCN axons crossing the optic chiasm between the lobula and the medulla was observed (Srahna, 2006).

Although the genetic data indicate that Dsh- and Rac-mediated signaling have sensitive and antagonistic effects on the JNK pathway, they do not establish whether the Dsh-Rac interaction modulates JNK's intrinsic activity. To test this, the amount of phosphorylated JNK relative to total JNK levels in fly brains was evaluated by Western blot analysis using phospho-JNK (P-JNK) and pan-JNK specific antibodies. Then it was determined if Dsh is indeed required for increased levels of JNK phosphorylation. Dsh1 mutant brains showed a 25% reduction in P-JNK consistent with a stimulatory role for Dsh on JNK signaling. The reduction caused by loss of Dsh function is reversed, when the amount of Rac is reduced by half, consistent with a negative effect of Rac on JNK signaling downstream of Dsh. These data support the conclusion that Dsh and Rac interact to regulate JNK signaling by modulating the phosphorylated active pool of JNK (Srahna, 2006).

Taken together, these data suggest that during brain development DCN axons extend under the influence of JNK signaling. A non-canonical Wnt signal acting via Fz and Dsh ensures that JNK signaling remains active by attenuating Rac activity. In contrast, activation of the FGFR activates Rac1 and suppresses JNK signaling. These data support a model whereby the balance of the Wnt and FGF signals is responsible for determining the number of DCN axons that stably cross the optic chiasm. To test this model, FGFR levels were reduced, using the dominant-negative btl transgene, in dsh1 heterozygous flies. It was found that simultaneous reduction of FGF and Wnt signaling restored the number of axons crossing the optic chiasm to almost wild-type levels (10.2, with 33% of the samples indistinguishable from wild-type, suggesting that the two signals in parallel, act to control the patterning of DCN axon connectivity (Srahna, 2006).

These data suggest the following model of DCN axon extension and retraction. DCN axons extend due to active JNK signal. These axons encounter Wnt5 and probably Wnt2 as well, resulting in activation of Disheveled. Disheveled, via its DEP domain, has a negative effect on the activity of the Rac GTPase, thus keeping JNK signaling active. After DCN axons cross the second optic chiasm they encounter a spatially regulated FGF/Branchless signal that activates the FGFR/Breathless pathway. Breathless in turn activates Rac, which inhibits JNK signaling in a subset of axons. These axons then retract back toward the lobula. The wide expression of the different components of these pathways and the modulation of JNK phosphorylation by Dsh and Rac in whole-head extracts strongly suggests that this model may apply to many neuronal types (Srahna, 2006).

Different Wnt signals act through the Frizzled and RYK receptors during Drosophila salivary gland migration

Guided cell migration is necessary for the proper function and development of many tissues, one of which is the Drosophila embryonic salivary gland. Two distinct Wnt signaling pathways regulate salivary gland migration. Early in migration, the salivary gland responds to a WNT4-Frizzled signal for proper positioning within the embryo. Disruption of this signal, through mutations in Wnt4, frizzled or frizzled 2, results in misguided salivary glands that curve ventrally. Furthermore, disruption of downstream components of the canonical Wnt pathway, such as dishevelled or Tcf, also results in ventrally curved salivary glands. Analysis of a second Wnt signal, which acts through the atypical Wnt receptor Derailed, indicates a requirement for Wnt5 signaling late in salivary gland migration. WNT5 is expressed in the central nervous system and acts as a repulsive signal, needed to keep the migrating salivary gland on course. The receptor for WNT5, Derailed, is expressed in the actively migrating tip of the salivary glands. In embryos mutant for derailed or Wnt5, salivary gland migration is disrupted; the tip of the gland migrates abnormally toward the central nervous system. These results suggest that both the Wnt4-frizzled pathway and a separate Wnt5-derailed pathway are needed for proper salivary gland migration (Harris, 2007).

Salivary gland migration can be separated into three phases. In the first phase, the salivary glands invaginate into the embryo at a 45° angle, moving dorsally until they reach the visceral mesoderm. fkh, RhoGEF2 and 18 wheeler have been shown to regulate apical constriction of the salivary gland cells during this invagination process. In addition, hkb and faint sausage are needed for proper positioning of the site of invagination. No guidance cues have been identified for this first phase of migration; it may be that the patterns of constriction and cell movements at the surface of the embryo are sufficient to direct the invaginating tube (Harris, 2007).

During the second phase of migration, as the salivary gland moves posteriorly within the embryo, two guidance cues, Netrin and Slit, guide salivary gland migration along the visceral mesoderm. Netrin, which is expressed in the CNS and the visceral mesoderm, works to maintain salivary gland positioning on the visceral mesoderm. At the same time, Slit acts as a repellent from the CNS to keep the salivary glands parallel to the CNS. A third guidance signal, WNT4, which acts through FZ or FZ2 receptors, is also required in the second phase of salivary gland migration. Loss of Wnt4, fz or fz2 in the embryo results in salivary glands that are curved in a ventromedial direction. This curving affects a large portion of the salivary gland and may result from the fact that the fz and fz2 receptors, in contrast to drl, are expressed throughout the salivary gland. Furthermore, dominant-negative transgenes that disrupt the function of DSH or TCF cause the same phenotype, suggesting that transcription induced by the canonical Wnt signaling pathway is needed to maintain the proper migratory path of the salivary glands on the circular visceral mesoderm (CVM). The migration along the CVM takes more than 2 hours for completion, which would leave adequate time for a transcriptional response (Harris, 2007).

Although Wnt4 and slit are both required for the second phase of migration, and their mutants show similar, though distinguishable, phenotypes, they are thought to act independently. While most slit-mutant embryos have medially curving salivary glands, embryos lacking Wnt4 have salivary glands that curved in a distinctly different, ventromedial, direction. Embryos doubly mutant for Wnt4 and slit show predominantly one or the other phenotype and neither phenotype increases in severity. These results suggest, though they do not prove, that Wnt4 and slit act in distinct pathways (Harris, 2007).

After the entire salivary gland has invaginated, migrated posteriorly within the embryo and lies parallel to the anteroposterior axis of the embryo, the distal ends of the salivary glands come into contact with the LVM. drl and Wnt5 are required for this late phase of salivary gland positioning. Loss of either drl in the salivary gland or Wnt5 in the CNS results in the distal tip of the salivary gland being misguided to a more ventromedial position. This change in the shape of the salivary gland is seen only after the salivary glands are no longer in contact with the CVM (after stage 13). Thus it is proposed that drl is required during the third phase of salivary gland migration, as the salivary gland detaches from the CVM and contacts the LVM (Harris, 2007).

The striking expression of drl at the tip of the salivary gland makes the leading cells uniquely different from the rest of the salivary gland cells. These cells project lamellipodia upon reaching the visceral mesoderm and beginning their posterior migration. They may act to both guide and pull the rest of the gland during migration. Cells at the tip of a migrating organ are frequently specialized to guide migration. For example, the coordinated migration of the tracheal branches in Drosophila is achieved by induction of distinct tracheal cell fates within the migrating tips. This is illustrated by the fact that FGF (Branchless) signaling becomes restricted to the tips of the tracheal branches soon after they begin to extend. The migration and growth of Drosophila Malpighian tubules provide another clear example of specialized cells needed at the tip of a migrating tissue. One cell is singled out to become the tip cell, which directs the growth of the Malpighian tubules as well as organizes the mitotic response and migration of the other cells forming each tubule. In other systems, such as Dictyostelium slugs, cells at the tip of a migrating group are required and solely able to guide migration. These results establish that the leading cells of the migrating salivary glands have a specialized role to play in proper salivary gland positioning. First they are required to initiate invagination within the embryo, then they actively participate in migration along the CVM, and finally they ensure that the distal tip of the gland will remain associated with the LVM at the end of the migratory phase (Harris, 2007).

Despite the fact that it has been firmly established that Wnt5 and drl are important for the final placement of salivary glands, the signaling pathways downstream are not well defined. Because salivary-gland expression of full-length drl can rescue the drl-mutant phenotype, but drl lacking the intracellular domain cannot, it is thought that the intracellular domain of DRL is important for signaling. Similarly, misexpression of full-length drl can misguide axons in the ventral nerve cord, but misexpression of drl lacking its intracellular domain cannot (Yoshikawa, 2003). The genetic interactions found in this study between drl and Src64 support recent findings suggesting that Src64 acts downstream of drl in the ventral nerve cord. In addition, the other Drosophila Src kinase, Src42, may be required at two stages, during salivary gland migration along the CVM and downstream of WNT5-DRL signaling as the gland moves onto the longitudinal visceral mesoderm (Harris, 2007).

Another intriguing finding of this study is the involvement of the two remaining Drosophila RYKs, Drl-2 and dnt, in salivary gland development. The phenotypes of Drl-2 and dnt mutants are less penetrant than drl mutants, but they are qualitatively very similar. Furthermore, embryos doubly heterozygous for drl and Drl-2 have salivary glands that resemble those seen in drl mutant embryos. These three RYKs appear to act in a partially redundant fashion in the salivary glands, since none of the single gene mutations leads to completely penetrant phenotypes. However, no increase was seen in penetrance of the drl phenotype in embryos lacking both drl and Drl-2. In addition, it was not possible to detect transcripts for either Drl-2 or dnt in the salivary gland. While it is possible that dnt and Drl-2 are expressed at very low levels in the salivary gland, they might be acting non-autonomously (Harris, 2007).

An interesting dilemma in understanding RYK signaling is how inactive kinases propagate a signal into the cell. Recent mammalian studies have suggested that RYKs may associate with another catalytically active receptor, such as FZ or EPH, at the membrane. In the mouse, the extracellular WIF domain of RYK interacts with FZD8, and it has been proposed that the two proteins may form a ternary complex with WNT1 to initiate signaling. However, data from flies and nematodes support the argument that DRL and its C. elegans homolog LIN-18 act independently of FZ. Genetic studies of cell specification in the nematode vulva suggest that LIN-18 acts in a parallel and separate pathway from the LIN-17/FZ receptor. Similarly, reduction of fz and fz2 gene activity in flies has no effect on a DRL misexpression phenotype in the ventral nerve cord (Yoshikawa, 2003). This study has shown that double mutants for the Wnt4 and Wnt5 ligands and for the fz and drl receptors both show strong enhancements in comparison to the single mutants, reinforcing the conclusion that these two ligands are activating different pathways. In addition, the functions of these two pathways can be separated by phenotype. The Wnt4-fz/fz2 phenotype becomes evident earlier and affects a larger portion of the salivary gland than the Wnt5-drl phenotype. Taken together, these results demonstrate that there are two independent Wnt pathways regulating salivary gland positioning. The early WNT4 signal appears to activate the canonical Wnt pathway, whereas there is a later requirement for WNT5 signaling through DRL and the Src kinases (Harris, 2007).

The planar polarity pathway promotes coordinated cell migration during Drosophila oogenesis

Cell migration is fundamental in both animal morphogenesis and disease. The migration of individual cells is relatively well-studied; however, in vivo, cells often remain joined by cell-cell junctions and migrate in cohesive groups. How such groups of cells coordinate their migration is poorly understood. The planar polarity pathway coordinates the polarity of non-migrating cells in epithelial sheets and is required for cell rearrangements during vertebrate morphogenesis. It is therefore a good candidate to play a role in the collective migration of groups of cells. Drosophila border cell migration is a well-characterised and genetically tractable model of collective cell migration, during which a group of about six to ten epithelial cells detaches from the anterior end of the developing egg chamber and migrates invasively towards the oocyte. The planar polarity pathway promotes this invasive migration, acting both in the migrating cells themselves and in the non-migratory polar follicle cells that they carry along. Disruption of planar polarity signalling causes abnormalities in actin-rich processes on the cell surface and leads to less-efficient migration. This is apparently due, in part, to a loss of regulation of Rho GTPase activity by the planar polarity receptor Frizzled, which itself becomes localised to the migratory edge of the border cells. It is concluded that, during collective cell migration, the planar polarity pathway can mediate communication between motile and non-motile cells, which enhances the efficiency of migration via the modulation of actin dynamics (Bastock, 2007).

This study used the Drosophila ovary to study the control of coordinated cell movements by the planar polarity pathway, taking advantage of its relative simplicity and the ability to precisely manipulate gene function in individual cell populations. Activity of the core polarity genes facilitates invasive migration of the border cell cluster through the nurse cells. Of particular interest is the observation that migration of the border cells is enhanced by planar polarity activity in the non-migratory epithelial polar follicle cells, suggesting a key role for interactions between migratory and non-migratory cell types (Bastock, 2007).

In the Drosophila wing, the planar polarity pathway regionalises cells via the formation of proximal and distal domains at the level of the adherens junctions. The distal domain contains Fz, which acts via the downstream factors Dsh and RhoA to ensure local production of a single actin-rich trichome, while, in the proximal domain, Stbm recruits factors that locally inhibit trichome formation. During border cell migration, the coordinated movement of the non-migratory polar follicle cells and the migratory border cells is achieved in part by the border cells retaining epithelial character in the region contacting the polar follicle cells, but also having an actin-rich partly mesenchymal migratory region. Taking these observations together, it is proposed that, in border cells, localised Fz in the migratory region and localised Stbm in the junctional region might promote the production of actin-rich structures, which, in turn, would increase the motility both of individual cells and the cluster as a whole (Bastock, 2007).

Mosaic analyses suggest a mechanism for how this localised Fz and Stbm activity is established within the border cells. Fz and Stbm mediate intercellular communication between the polar cells and the border cells via the production of junctional complexes. Because contact with an Fz-expressing polar cell enhances the migration of border cells, it is surmised that Fz in each polar cell interacts with Stbm in the contacting border cell. Junctionally localised Stbm in the border cell can then act as a cue to indirectly promote actin-rich protrusion formation in the migratory region, at least in part via the localisation of Fz (Bastock, 2007).

Although the planar polarity pathway has been known for some years to promote cell rearrangements during vertebrate gastrulation, surprisingly little is understood about its mechanisms of action in cell movement and the particular roles of this pathway in cell-cell communication. This study has demonstrated that Fz/Stbm-mediated intercellular communication can enhance the invasive migration of a group of cells. Migration of groups of cells, sometimes including both motile and non-motile types, is important for many processes in animal morphogenesis and in disease processes, such as cancer metastasis. This work provides evidence that planar polarity pathway function could be generally important in coordinated cell migration, providing a mechanism by which cells within a group can communicate and establish the proper regional production of actin structures required for efficient movement (Bastock, 2007).

Combinatorial signaling by the Frizzled/PCP and Egfr pathways during planar cell polarity establishment in the Drosophila eye

Frizzled (Fz)/PCP signaling regulates planar, vectorial orientation of cells or groups of cells within whole tissues. Although Fz/PCP signaling has been analyzed in several contexts, little is known about nuclear events acting downstream of Fz/PCP signaling in the R3/R4 cell fate decision in the Drosophila eye or in other contexts. This study demonstrates a specific requirement for Egfr-signaling and the transcription factors Fos (AP-1), Yan and Pnt in PCP dependent R3/R4 specification. Loss and gain-of-function assays suggest that the transcription factors integrate input from Fz/PCP and Egfr-signaling and that the ETS factors Pnt and Yan cooperate with Fos (and Jun) in the PCP-specific R3/R4 determination. The data indicate that Fos (either downstream of Fz/PCP signaling or parallel to it) and Yan are required in R3 to specify its fate (Fos) or inhibit R4 fate (Yan) and that Egfr-signaling is required in R4 via Pnt for its fate specification. Taken together with previous work establishing a Notch-dependent Su(H) function in R4, it is concluded that Fos, Yan, Pnt, and Su(H) integrate Egfr, Fz, and Notch signaling input in R3 or R4 to establish cell fate and ommatidial polarity (Weber, 2008).

Previous studies established that Fz is required cell-autonomously for R3 fate induction. The current analyses of kay/fos LOF alleles indicate that Fos is also required cell-autonomously in R3 for its fate determination. When overexpressed, Fos also acts like Fz in R3/R4 photoreceptors at the time of PCP establishment, with the cell of the pair that has higher Fos levels adopting the R3 fate. Based on its requirement in R3 and genetic interactions, Fos could act as a nuclear effector of Fz/PCP signaling. This is supported by the observation that it is able to suppress sev-dsh induced PCP defects; the genetic data can however not rule out that Fos could act in parallel to Fz/Dsh-PCP signaling). The subtle differences observed between fz and kay/fos LOF requirements (in fz R3/R4 mosaics the wild-type cell adopts the R3 fate often causing chirality inversions, while in kay/fos mosaic pairs with a mutant R3 the pair often adopts symmetrical R4/R4 appearance) is likely due either to the hypomorphic nature of the kay/fos alleles that had to be used in the analysis or potential redundancy with jun (Weber, 2008).

In addition to the positive Fos signaling input, R3 specification also requires the repressor function of Yan, with Yan inhibiting R4 fate in the R3 precursor. This is evident by the cellular requirement of Yan and highlighted by the increased defects in a kay/fos and yan double mutant scenario, where both aspects are partially impaired causing frequent R3/R4 fate decision defects. The dominant enhancement of kay2 by yan LOF suggests that keeping the R4 fate off in R3 precursors is as important as inducing the R3 fate (Weber, 2008).

Previous work has demonstrated that Fz/PCP signaling leads to Dl and neur upregulation in R3, activating Notch signaling in the neighboring R4 precursor. This study shows that Egfr-signaling is also specifically required for R4 fate determination. The ETS factors Yan and Pnt are nuclear effectors of Egfr-signaling in many contexts including photoreceptor induction, and the data indicate that they act also in R3/R4 determination. Egfr-signaling leads to an inactivation of Yan and an activation of Pnt through their phosphorylation by the Rl/Erk MAPK. As Yan represses the R4 fate it needs to get inactivated in the R4 precursor by Egfr-signaling and conversely Pnt is activated in R4. Together with the Notch-Su(H) activity this leads to R4 fate induction. Thus, for R3 determination Fz/PCP signaling and its nuclear effectors Fos (and Jun) are sufficient, along with Yan mediated repression of the R4 fate in R3 precursors. R4 fate determination, on the other hand, requires the joint activity of two pathways, Notch and Egfr-signaling and their nuclear effectors. A similar Egfr-Notch cooperation is observed in R7 induction and in cone cells (Weber, 2008).

These data support a complex interaction scenario between Fz/PCP, Notch, and Egfr-signaling in R3/R4 fate determination. Whereas the Notch-Su(H) activation in R4 depends on Fz/PCP signaling in the R3 precursor, the Fz/PCP and Egfr-signaling pathways require a fine balance. This is reflected by their genetic interactions, both at the level of the receptors fz and Egfr and their nuclear effectors Fos/Jun and the ETS factors Pnt and Yan, suggesting a cooperative involvement between the Fz/PCP and Egfr pathways (Weber, 2008).

The nuclear Egfr-signaling response is very likely mediated by Pnt in R4. Although this could not be addressed in pnt LOF clones due to the non-autonomous defects, which are caused by feedback loop requirements in which Pnt participates. The sufficiency experiments fully support a cell-autonomous requirement of Pnt in R4 to specify R4 fate, consistent with the Egfr requirement (Weber, 2008).

In summary, the behavior of the nuclear effectors of the respective signaling pathways involved in R3/R4 specification reflects the combinatorial nature of the signaling pathway input into the R3 and R4 fates (Weber, 2008).

Although in the embryo Fos and Jun need to act as heterodimeric partners in a non-redundant manner, in imaginal discs the scenario is more complicated. Whereas jun mutant clones display only mild phenotypes and do not affect proliferation/survival, strong kay/fos LOF alleles show severe defects, suggesting that kay/fos is the main AP-1 component acting in imaginal discs. This is supported by recent studies on the role of Fos in cell cycle regulation and proliferation (Hyun, 2006). Nevertheless, the double mutant combination of kay and jun revealed a requirement of both as no kay/fos, jun double mutant cells are recovered, suggesting a partially redundant function of kay/fos and jun in imaginal discs (Weber, 2008).

The specific role of the possible distinct heterodimers between the different Fos isoforms and Jun, or the different Fos isoforms themselves, could be very complex. This complexity is also evident in the fact that overexpression of a dominant-negative Fos protein form or a single wild-type isoform (transcript RA, according to Flybase) causes similar phenotypic defects (e.g. in the eye or in thorax closure). Future experiments will have to address which of the Fos isoforms is required in which context and if and how they interact with Jun (Weber, 2008).

Wnt/PCP proteins regulate stereotyped axon branch extension in Drosophila

Branching morphology is a hallmark feature of axons and dendrites and is essential for neuronal connectivity. To understand how this develops, the stereotyped pattern of Drosophila mushroom body (MB) neurons, which have single axons branches that extend dorsally and medially, was analyzed. Components of the Wnt/Planar Cell Polarity (PCP) pathway were found to control MB axon branching. frizzled mutant animals showed a predominant loss of dorsal branch extension, whereas strabismus (also known as Van Gogh) mutants preferentially lost medial branches. Further results suggest that Frizzled and Strabismus act independently. Nonetheless, branching fates are determined by complex Wnt/PCP interactions, including interactions with Dishevelled and Prickle that function in a context-dependent manner. Branching decisions are MB-autonomous but non-cell-autonomous as mutant and non-mutant neurons regulate these decisions collectively. Wnt/PCP components do not need to be asymmetrically localized to distinct branches to execute branching functions. However, Prickle axonal localization depends on Frizzled and Strabismus (Ng, 2012).

Branching morphogenesis is a key feature in axons and dendrites. This study shows that Wnt/PCP proteins, known for their roles in tissue polarity, are involved in two aspects of axon branch formation: stereotyped extension and collective decision-making. Some PCP mutants have a preference for dorsal or medial branching phenotypes. As the genetic results suggest that Fz and Stbm functions are separable, this is distinct from some tissue polarity models that have been presented for Drosophila patterning. The results suggest that PCP proteins have complex cross-regulatory effects that include the spatial regulation of Pk. A balance of their functions is key to stereotyped branching and to axon growth and guidance (Ng, 2012).

An independent study recently reported similar results for Wnt/PCP components in MB axon branching, guidance and growth (Shimizu, 2011). When comparing the results with those described in this study, most of the data are consistent. Slight differences, such as the immunohistochemistry results and the role of Prickle, might be a result of differences in preparative methods, stages at which animals were analyzed and the genetic background of the animals used (Ng, 2012).

Branched axons and dendrites tend to follow a bifurcation motif. In MB axons, this corresponds to a single dorsal and a single medial projection. This study shows that PCP proteins control the extension of axon branches, rather than the initial branch formation. The working model is that growing axons can initially reach the axon peduncle and bifurcate independently of the Wnt/PCP pathway. Once the nascent branches have formed, Fz and Stbm direct the stereotyped extension of branches. Fz is the major determinant of dorsal projections. By contrast, Stbm predominantly controls medial projections. The results suggest that these are partly redundant systems; dorsal and medial projections also require minor contributions from Stbm and Fz, respectively. In the absence of both, branching errors are increased and randomized. In some cases, both extensions are lost. Axon branching relies mainly on mutual cooperation between Fz and Stbm. In some cases, antagonism between Stbm and Fz exists, although there is no evidence of Stbm inhibiting Fz signals. These results are reminiscent of observations in Drosophila showing that Fz and Stbm can exhibit redundant and cooperative PCP functions in the wing as well as in sensory organ precursor cell fate specification (Ng, 2012).

MB axons follow a branching mode that reflects collective behavior. The initial results suggest that PCP mutant phenotypes are recessive. Next to wild-type axons, PCP mutant axons tend to adopt a wild-type branched pattern. However, with the GFPstbm rescue paradigm, the results together suggest that it is the collective mix of mutant and non-mutant neurons that determines whether an axon branches or not. How is this regulated? One possibility is that as MB axons are fasciculated, cell-cell contacts can regulate collective branching behavior. Although axons are long and thin, it is unlikely that 1000 branched αβ neurons in each hemisphere are in contact with each other. It is more likely that, as in other Drosophila tissues, it is the behavior of cells that are widely interlinked together that is controlled by PCP factors. The non-autonomous results presented in this study differ significantly from previous PCP studies. Previous results in Drosophila wing and eye show that PCP mutations result in both cell-autonomous and non-autonomous tissue polarity defects. In this study PCP branch functions are not fully cell-autonomous. Furthermore, if domineering non-autonomy exists, it does not strongly affect wild-type axon branching (Ng, 2012).

Despite the uncertainty about its role, asymmetric organization of PCP proteins is a key feature in tissue polarity. However, there are variations to this theme. For example, although Fz and Stbm are often on opposite poles, in mouse eye lens fiber cells they appear in the same tissue pole. In zebrafish and ascidian models in which PCP signals control the convergence and extension of cells in the developing notocord, PCP protein localizations are variable. These and other observations have led to the suggestion that in motile cells, PCP proteins might be dynamically regulated. In short, despite PCP mutants having different preferences in dorsal or medial branch defects, no obvious PCP protein asymmetry can be found between axon branches. One possibility is that these proteins are regulated dynamically, therefore such differences are not readily observable in fixed tissue. Another possibility is that distinct branching fates might arise from differences in PCP effector functions between branches (Ng, 2012).

The observation that Pk axonal branch levels are reduced in fz and stbm mutants suggests that Pk regulation forms part of PCP branch functions. This is likely to be post-translational and axonal-specific as Pk localization in MB dendrites is unaffected by fz or stbm. This is similar to Drosophila wing patterning for which Fz loss also results in a cell-autonomous reduction of Pk. However, in the wing, Dsh, Fz and Stbm also regulate the intercellular accumulation of Pk at the cell cortex, along the pole vertices. Also, in contrast to some wing models, this study shows that Fz and Stbm branch functions can still occur independently of Pk (Ng, 2012).

Several results suggest that PCP proteins function early during axonal development. First, Pk expression is higher in the αβ core region, where newborn developing MB axons are located. Further evidence from the stbm rescue studies shows that GFPstbm expression in the αβ core alone is sufficient to rescue axon branching. Also, branching defects were apparent in PCP mutants at early stages of development. Lastly, Pk and Dsh αβ localization can also be detected at these stages (Ng, 2012).

How Wnt/PCP proteins interact has been studied extensively in Drosophila. This study shows that the MB axon phenotypes reveal some novel features of PCP signaling interactions. As already mentioned, Fz and Stbm play partly redundant, parallel roles. Although Pk controls axon branching and is regulated by Fz and Stbm, it does not account for all Fz and Stbm functions. This is also evident from other PCP studies. However, Pk does act additively to Fz and Stbm and has convergent as well as antagonistic functions with Dsh. Also, this study has shown that Dsh does not simply transduce Fz or Stbm signals. This was also reported in some studies. The working model is that both Fz and Stbm are likely to function in part through Dsh and Pk. Based on further experiments, their effects are likely to be complex and context-dependent on the activity levels of other PCP proteins (Ng, 2012).

Previous reports suggest that these factors interact physically with each other. There is evidence that the extracellular regions of Stbm and Fz can physically bind to each other, although this is debated. Physical associations between Stbm-Dsh and Fz-Dsh have been documented in vertebrate and fly proteins. Apart from Stbm, Pk also interacts physically with Dsh. Recent reports suggest that mouse Stbm (Vangl2) and Dsh (Dvl1) interact antagonistically to regulate Frizzled3 (Fzd3) phosphorylation. Whether these interactions affect axon branching is currently unknown. Nonetheless, novel Wnt/PCP interactions are involved in MB axon patterning (Ng, 2012).

In conclusion, this study demonstrates that Wnt/PCP proteins are key regulators in the stereotyped growth of axon branches. Given that branched growth is intrinsic to neuronal morphogenesis, Wnt/PCP components and their reported interactions might also be involved synapse and dendrite formation. Such highly cross-regulated programs are probably essential for the intricate patterning of the nervous system (Ng, 2012).

dachsous and frizzled contribute separately to planar polarity in the Drosophila ventral epidermis.

Cells that comprise tissues often need to coordinate cytoskeletal events to execute morphogenesis properly. For epithelial tissues, some of that coordination is accomplished by polarization of the cells within the plane of the epithelium. Two groups of genes--the Dachsous (Ds) and Frizzled (Fz) systems--play key roles in the establishment and maintenance of such polarity. There has been great progress in uncovering the how these genes work together to produce planar polarity, yet fundamental questions remain unanswered. The Drosophila larval ventral epidermis has been studied to begin to address several of these questions. ds and fz are shown to contribute independently to polarity, and they do so over spatially distinct domains. Furthermore, it was found that the requirement for the Ds system changes as field size increases. Lastly, it was found that Ds and its putative receptor Fat (Ft) are enriched in distinct patterns in the epithelium during embryonic development (Donoughe, 2011).

In early embryos, the body axis is subdivided into parasegments, each of which is further subdivided into two domains. One half of the epithelial cells will secrete smooth cuticle and the other half will form cuticular protrusions called denticles (the denticle field). The denticle field pattern is the product of a series of distinct polarized events. First, cells align into columns as a consequence of the reorganization of select cell interfaces. Second, one to three F-actin bundles protrude from the posterior edge of each cell. Third, the F-actin bundles guide the secretion of extracellular matrix (cuticle) such that denticles take on their final tapered orientation and hooked shapes. The result is that each column of denticles corresponds to a single column of underlying cells. This study has taken advantage of this polarized pattern to investigate the roles of ds, ft and fz in establishing this planar polarity (Donoughe, 2011).

With each molt, a growing larva secretes a new cuticle that is patterned on the underlying epidermis. Since there are no major cell rearrangements nor any increase in cell number during larval growth, cells of this epithelium maintain their specific fates and relative positions. Thus, the denticle pattern is resynthesized for each successive cuticle, where the columns of protruding denticles remain intact until the next molt, enabling the crawling larvae to grip the substrate during locomotion (Donoughe, 2011).

This study addresses long-standing questions in the planar cell polarity (PCP) field: (1) how do Fz and the members of the Ds system each contribute to planar polarity in an epithelium and (2) how do Ds and Ft influence the polarized placement of F-actin protrusions (Donoughe, 2011)?

This study has elucidated the contributions of several key polarity genes in the larval ventral epidermis. The genes in the Ds system are essential for proper polarity in this tissue. Notably, the Ds extracellular domain is able to reorient adjacent cells even when they are null for ds. The Fz protein operates largely redundantly and in parallel to the Ds system, and appears to contribute more in some columns than others. As field size increases, it is likely that Fz is less able to polarize the tissue on its own. By contrast, the Ds system is able to polarize the tissue equally well at small and large field sizes. Finally, it was found that in embryos, Ds and Ft are enriched in the posterior half of each denticle field. This correlates with the domain of the embryonic denticle field where actin protrusion placement defects appear in ds M-Z- embryos (Donoughe, 2011).

Several observations are in line with what is understood from other tissues. First, the polarity disruptions in ds- or ft- single mutants are comparable in severity to those observed in ds- ft- double mutants. This confirms that Ds and Ft act within the same process to polarize tissues. Second, in the adult abdomen, an experimentally induced high point of Ds extracellular domain expression causes an adjacent cell to reorient its polarity toward that high point (see Ds extracellular domain can reorient neighboring denticle columns). Likewise, overexpression of the Ds extracellular domain in one cell column of an otherwise wild-type larva causes the flanking cell columns to reorient toward this (presumed) enhanced source of Ds. By repeating this experiment in the ds- mutant, any potentially confounding contributions from the superimposed distribution of endogenous Ds were avoided. Therefore, it can be concluded that cells polarize toward high levels of Ds. Whether this is the case during normal patterning is more difficult to address (Donoughe, 2011).

Finally, it was found that gain-of-function effects are propagated farther than just the adjacent cell. Thus, in a wild-type background, excess Ds in column 1 caused reorientation in columns 2 and 3. This implies that the signal was received in column 2 (resulting in altered polarity there), and then a polarizing effect was propagated to column 3. When such overexpression was repeated in a ds- background, however, column 2 reoriented whereas column 3 largely did not. This demonstrates that Ds is not required for a cell to respond to a Ds polarity signal, but it is important in propagating that signal onward. Altogether, these findings support the hypothesis that Ds and Ft work together to send, implement and propagate a polarity signal (Donoughe, 2011).

A central focus of ongoing research is to determine how the Fz and Ds systems each contribute to the establishment and maintenance of planar polarity. In both the Drosophila eye and wing it appears that the Ds system provides a directional cue that is amplified and implemented by the Fz system. In the abdomen, by contrast, the Ds system can polarize in the absence of Fz and Stan, both of which are essential for the non-cell-autonomous effects of the Fz system. The current findings make it clear that for the larval denticle field, the Fz protein acts in a way that is inconsistent with its proposed role downstream of the Ds system. However, this observation still leaves room for the possibility that Ds-Ft engages downstream components within the Fz system (Donoughe, 2011).

The larval epidermis is unique in that the relative requirements for the Ds and Fz systems differ in different domains. The most obvious example of this is that when the Ds system is removed, polarity is completely removed in some columns (e.g., columns 0 and 4) but at least some polarity is still present in others (e.g., columns 1, 2, 3, 5, 6). Thus, it appears that the Fz system (still intact) is acting in those columns to impart polarity, suggesting that the two systems have independent and redundant inputs to polarity (Donoughe, 2011).

It was also demonstrated that Ds extracellular domain overexpression is able to reorient adjacent columns in an fz null background, and this signal is propagated onward. This shows that the Ds system can send, receive and propagate polarity information without contribution from the Fz protein. It remains possible that even when Fz-dependent intercellular signaling is absent, intracellular components of the Fz-system, such as Dsh, act in implementing the Ds signal. This function of Dsh would have to be unaffected in dsh[1] MZ mutants, as the polarity of dsh[1] MZ and dsh[1] MZ ds- larvae appear similar to that of fz- and ds- fz- larvae, respectively. Testing for Ds-mediated polarity in dsh null cells would be the true test of this hypothesis, but is precluded by the essential role of dsh in canonical Wnt signaling (Donoughe, 2011).

If, however, the Ds system operates independently of the Fz system, this would have significant ramifications for understanding of the molecular mechanisms that must be engaged downstream of each polarity system. The two systems must eventually converge at the point when cells create the oriented read-out (in this case, denticle formation). It is possible that the common polarity effectors might be far downstream of the initial effects in signaled cells. Given that the Fz and Ft receptors are so dissimilar from a molecular standpoint, their immediate effectors are likely to be distinct. Only by identifying the proteins that interact with Ft to implement Ds system polarity will it be possible to determine whether these effectors intersect downstream components of the Fz system or act independently on the polarity read-out (Donoughe, 2011).

Another observation that requires explanation is that the Ds and Fz systems seem to operate serially in some contexts (e.g. in the eye or wing) but in parallel in others (e.g. in the abdomen or the larval epidermis). Ds system-mediated microtubule (MT) orientation has been suggested as one mechanism by which the Ds system could feed into the Fz system. When MTs are oriented along the axis of polarity of wing cells, MT-mediated polarized transport brings Fz to the cell membrane, and it was recently shown that the maintenance of the correct MT orientation is Ds dependent. In the embryonic ventral epidermis, however, MTs are oriented perpendicular to the axis of planar polarity, at least at steady state. Therefore, unless careful imaging uncovers a minor, posteriorly polarized and Ds-dependent MT track, it seems unlikely that the Ds system is operating in the ventral epidermis in the manner proposed for wing polarity. This could explain why the Ds system only functions independently of Fz protein in the denticle field (Donoughe, 2011).

In ds- fz- larvae, all columns were largely disordered, but the flanking cell columns exhibited a slight, yet statistically significant, tendency toward reversed polarity. It is difficult to explain why there is residual polarity rather than randomization. Although the ds and fz alleles that were used are nulls, fz2 cannot be additionally removed owing to its essential role in canonical Wnt signaling. Thus, it is possible that Fz2-dependent polarization makes some contribution in the larval epidermis, although Fz2 has not as yet been implicated in PCP in any tissue. Alternatively, even if fz were the only Fz system receptor active for PCP, some latent activation of downstream components of the Fz system could, in principle, be responsible for imparting this subtle but polarized output (Donoughe, 2011).

An alternative explanation for the residual polarization in ds- fz- mutants is that there is an underlying bias in the tissue that is ordinarily masked in the presence of Ds or Fz proteins, but uncovered when both are removed. Since the residual orientation in double mutants tends to be directed away from the smooth field, perhaps that domain is somehow responsible for the latent polarity. Alternatively, the orientation signal might derive from within the denticle field. For instance, the 4-5 column interface is a boundary for Notch and EGFR signaling. Perhaps a low-level orientation signal emanates from that position (Donoughe, 2011).

This work also suggests that the Ds and Fz systems have different capacities to adjust to changes in field size. Current models for creating planar polarity begin with a gradient across the field of unpolarized tissue. A subtle bias is presumably then established within each cell across the field, as cells compare the level of the polarizing gradient they detect with that detected by their neighbors. This bias is then reinforced in each cell through a feedback mechanism, converting it into a sharp intracellular gradient of effector protein distribution (Axelrod, 2009). At those initial stages, when a given cell compares the level it perceives with that of adjacent cells, the magnitude of the difference under comparison should be influenced by the size of the field: as field size increases, the contrast perceived by adjacent cells decreases. Correspondingly, any comparison mechanism will be challenged as field size increases (Donoughe, 2011).

The larval epidermis presents such a challenge to the polarizing systems as tremendous growth occurs across the field between each larval molt. This study succeeded in analyzing the effects on the Fz system as field size increased by examining ds null animals at each molt. At small field size (i.e. first instar), polarity defects are rare; however, at large field size (i.e., third instar, five times larger), the disruption to polarity is dramatic. This suggests that the Fz system loses potency as field size increases. By contrast, the Ds system did not appear to be affected, as there are only rare defects in fz null animals at first or third instar. Since the change in field size through the larval instars occurs in the absence of cell division, it will be of interest to explore what other parameters of cell growth affect the Fz system in this tissue (Donoughe, 2011).

Note also that this work demonstrates that denticle field polarity can change over the course of larval growth. This supports the recent finding that third instar polarity is not determined at the embryonic stage. Together, these findings strongly imply that planar polarity in the larval epidermis is not permanently set, but rather requires input throughout larval growth (Donoughe, 2011).

The ventral epidermis also provides the opportunity to study how the two polarity systems influence distinct polarized outputs from the same tissue. Cell alignment and denticle orientation were largely unaffected in ds M-Z- embryos/first instar larvae, but there were F-actin protrusion placement defects in cell columns 3 through 5. This result is compelling, as the domain affected matched the region of peak Ds and Ft accumulation. In fz M-Z- and dsh[1] MZ backgrounds, there are subtle column 1 and 2 defects in F-actin protrusion placement. It is intriguing that the embryonic protrusion placement defects appear in complementary patterns for the Fz system as compared with the Ds system; this suggests that in embryos, as in larvae, the two systems function mainly in spatially distinct domains (Donoughe, 2011).

In several tissues, protein distributions have provided a window into the mechanism of polarization. However, in the embryonic epidermis, this analysis so far has not been suggestive. As neither Ds nor Ft showed an obvious bias toward particular interfaces around a given cell, it is not immediately apparent how these accumulation patterns might be related to proposed Ds-Ft dimer distributions or to the polarity of the tissue. It is of course possible that the protein accumulations would be more suggestive if one could analyze them during the larval molts, but this cannot presently be done (Donoughe, 2011).

In this context, it is worth noting that the endogenous distributions of Fz system components have not yet been determined in the ventral epidermis. Staining for Fz-GFP and Dsh-GFP, however, reveals a difference in their enrichments as compared with Ds and Ft: both Fz system members are strongly enriched along cell interfaces that separate cell columns and are depleted from interfaces between cells within the same column. Whether these putative enrichments are necessary for polarity in this tissue remains to be tested (Donoughe, 2011).

Effects of mutation - Return: part 1/2

frizzled: Biological Overview | Evolutionary Homologs | Regulation | Developmental Biology | References

date revised: 25 October 98  

Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.

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