SEARCH

SEARCH BY CITATION

Keywords:

  • planar polarity;
  • PCP;
  • Drosophila;
  • Fat;
  • Dachsous

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. THE FT/DS SYSTEM AND PLANAR POLARITY
  5. THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?
  6. DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION
  7. INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING
  8. ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?
  9. PERSPECTIVES
  10. Acknowledgements
  11. REFERENCES

Planar polarity is generated through the activity of two groups of proteins, the “core” system and the Fat (Ft)/Dachsous (Ds) system. Although both are conserved from insects to mammals, vertebrate studies into planar polarity have primarily focussed on core planar polarity proteins and have only recently branched into the study of the Ft/Ds system. In Drosophila, however, years of detailed analysis have started to elucidate some of the mechanisms by which Ft/Ds signalling might set up polarity across a tissue, and how this may impact upon core protein-mediated planar polarity. In this review, we discuss the major findings, models, and controversies that have emerged from Drosophila research into the Ft/Ds system, and indicate some areas for further investigation. Developmental Dynamics 241:27–39, 2012. © 2011 Wiley Periodicals, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. THE FT/DS SYSTEM AND PLANAR POLARITY
  5. THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?
  6. DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION
  7. INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING
  8. ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?
  9. PERSPECTIVES
  10. Acknowledgements
  11. REFERENCES

Cell polarisation is essential for the development and functioning of tissues. In addition to exhibiting apico-basal polarity, epithelial cells possess a second type of polarity, known as planar polarity, which runs perpendicular to the apico-basal axis. Planar polarity is also seen occasionally in cell types that have no apico-basal polarity, such as some layers of mesenchymal cells. Until about 10 years ago when gene functions implicated in planar polarity were shown to be essential for gastrulation in Xenopus and zebrafish (Heisenberg et al., 2000; Wallingford et al., 2000), the field of planar polarity was a small corner of Drosophila developmental biology. However, it has become increasingly clear that the genetic control of planar polarity establishment is highly conserved between flies and vertebrates, and that planar polarity is essential for the correct anatomy and function of many tissues and organs, something that evidently has implications for human development and health.

Studies in Xenopus, mouse, zebrafish, and chick have demonstrated the importance of planar polarity in regulating multiple aspects of vertebrate development including orientation of cilia that regulate fluid movement in organs such as the kidney, positioning of stereocilia in the inner ear, axon guidance in the nervous system, and convergent extension movements during neural closure (reviewed in Wang and Nathans, 2007; McNeill, 2010; Goodrich and Strutt, 2011). Mutations in core planar polarity genes or downstream effectors have also been found to contribute to the aetiology of human birth defects such as spina bifida, and are implicated in diseases including polycystic kidney disease and various types of cancer (for more detailed reviews see Simons and Mlodzik, 2008; McNeill, 2009; Wang, 2009). However, despite the large number of recent publications examining the roles of planar polarity in vertebrates, the redundancy of polarity genes and complexity of polarised vertebrate tissues, together with long generation times and technical limitations mean that Drosophila remains the best model system for answering many of the outstanding mechanistic questions.

The general properties of planar polarity were discovered in the middle of the last century by tissue transplantation experiments in invertebrates such as the moth Galleria or the blood-sucking insect Rhodnius (reviewed in detail in Strutt, 2009). However, Drosophila tissues, in particular the wing and the eye, were rapidly adopted for their ease of use in determining the genetic control of planar polarity establishment (for more detailed reviews see Lawrence et al., 2007; Axelrod, 2009; Strutt, 2009; Goodrich and Strutt, 2011). The Drosophila wing develops from a single epithelial layer, and polarity is easily observable in pupal and adult wings by the presence of small wing hairs, also known as trichomes, which form at about 32 hr after pupa formation (32h APF) from the distal side of each cell (Fig. 1A,C; Wong and Adler, 1993). Planar polarity is also required for the alignment of cuticle ridges, which run along the antero-posterior axis in the anterior of the wing and the proximo-distal axis in the posterior of the wing (Doyle et al., 2008). The eye is slightly more complex as it is composed of groups of several different cell types, making up chiral units called ommatidia. Planar polarity in the eye is manifest in the breaking of symmetry in each ommatidium by the fate specification of two of the presumptive photoreceptors, R3 and R4 (Tomlinson and Ready, 1987; Zheng et al., 1995). Within each R3/R4 pair, the cell that is closest to the equator (the midline between the dorsal and ventral halves of the eye) takes on R3 identity, which determines both the chirality of the ommatidium and the direction in which it rotates (Fig. 1B,D). The absence of planar polarity cues in these tissues results in a “swirly” pattern of trichome orientation in the wing and defects in ommatidial chirality and rotation in the eye (Fig. 1E–H; Gubb and García-Bellido, 1982; Zheng et al., 1995).

thumbnail image

Figure 1. Planar polarity phenotypes in Drosophila tissues. Wings are oriented with proximal (prox.) to the left, distal (dist.) to the right, anterior (A) upwards, and posterior (P) downwards. Eyes are oriented with anterior (A) to the left, posterior (P) to the right, dorsal (D) upwards, and ventral (V) downwards. A: Trichome polarity in the adult wing points distally (green arrows), whereas cuticle ridges (red lines) are aligned along the antero-posterior axis in the anterior wing and along the proximo-distal axis in the posterior wing. B: Eye development in Drosophila. The eye develops through a wave of differentiation that moves from posterior to anterior (blue curved line). As ommatidia are specified, the R3/R4 fate choice depends upon which photoreceptor is closer to the equator; the closest of each pair becoming R3 (green) and the other R4 (orange). This process requires Fz in R3 and Stbm in R4. Using this information, ommatidial clusters undergo rotation and form their characteristic chiral shapes. C: Wild-type wings produce a distally pointing trichome from each cell. D: Wild-type eyes possess an ommatidial array that has uniform chirality on each side of the equator. E: In ds mutant wings, coordination of cell polarity is lost and trichomes form a swirling pattern. F: Chirality is randomised in ds mutant eyes. G,H: Comparison of defects in ommatidial chirality and rotation in mutant clones of fz (G) and ft (H). G: Inside fz clones, ommatidia show defects in rotation and chirality. Non-autonomous chirality defects are also seen at the polar edge of the clone (circled). H: Inside ft clones, the chirality of ommatidia is randomised, but there are no rotation defects. Non-autonomy is seen at the polar edge of the clone (circled). Images (C–F) courtesy of A. Brittle.

Download figure to PowerPoint

Drosophila genetics has shown that there are two major groups of proteins that mediate planar polarity: the core system and the Ft/Ds system, plus a series of effector proteins that are often tissue-specific, downstream transducers of polarity, and will not be mentioned further in this review (but for more details see Jenny, 2010; Goodrich and Strutt, 2011). As their name suggests, core planar polarity proteins have a central role in generating planar polarity in most tissues, and have been particularly well studied in the eye and wing primordia where they accumulate asymmetrically at the level of epithelial adherens junctions in an interdependent manner (Fig. 2). In the Drosophila wing, for example, proteins localise either distally (Frizzled [Fz], Dishevelled [Dsh], and Diego [Dgo], Axelrod, 2001; Feiguin et al., 2001; Strutt, 2001; Das et al., 2004), proximally (Strabismus [Stbm] and Prickle [Pk], Tree et al., 2002; Bastock et al., 2003), or in the case of Flamingo (Fmi) to both sides of the cell (Usui et al., 1999), and a similar asymmetry is seen in the eye at the R3/R4 boundary (Strutt et al., 2002). Core proteins are not distributed evenly along the proximal/distal cortex, but become stably clustered into asymmetric complexes that persist over a long period of time and reach peak asymmetry in the wing just before trichomes form (Classen et al., 2005; Aigouy et al., 2010; Strutt et al., 2011).

thumbnail image

Figure 2. The localisation of core planar polarity proteins. This cartoon represents a lateral view of a single junction (A) and an apical view of a field of cells (B) within the Drosophila wing. Proximal is to the left. Fmi (red) is localised to apical junctions on both proximal and distal sides of the cell, whilst the transmembrane proteins Fz (green) and Stbm (orange) are localised apically on distal and proximal membranes, respectively. The cytoplasmic proteins Dsh (navy) and Dgo (pink) associate with the complex distally, whilst Pk (blue) is recruited proximally (A). Trichomes form at the distal side of the cell in response to core protein localisation (B).

Download figure to PowerPoint

Although it seems increasingly likely that core protein interactions and feedback loops can account for local coordination of polarity and amplification of protein asymmetry (reviewed in Strutt and Strutt, 2009), the mechanism by which core protein asymmetry is globally aligned within a tissue is still poorly understood. Given that Fz proteins are well-characterised receptors for Wnt ligands (Bhanot et al., 1996), and Dsh is a member of the canonical Wnt signalling pathway (Klingensmith et al., 1994), a gradient of Fz activity in response to Wnt signalling was originally suggested to provide the missing long-range polarity cue (Adler et al., 1997). However, although several Wnts have been implicated in planar polarity signalling in vertebrates (Heisenberg et al., 2000; Tada and Smith, 2000; Qian et al., 2007), there is little evidence that any of them play a role in Drosophila (Lawrence et al., 2002; Chen et al., 2008).

An alternative mechanism for aligning core planar polarity with the body axes has been proposed from analysis of the Ft/Ds system. The Ft, Ds, and Four-jointed (Fj) proteins play an important role in controlling tissue growth through regulation of the Hippo (Hpo)/Warts (Wts) signalling cascade, in addition to their function in planar polarity specification (reviewed in Strutt and Strutt, 2005; Reddy and Irvine, 2008; Sopko and McNeill, 2009). As with the core system, the role of Ft/Ds in planar polarity is conserved in vertebrates (Saburi et al., 2008; Mao et al., 2011a), although studies are still in their infancy. Work in some Drosophila tissues has suggested a role upstream of the core (Adler et al., 1998; Rawls et al., 2002; Strutt and Strutt, 2002; Yang et al., 2002; Ma et al., 2003), but there is controversy surrounding the nature of the interaction (for more details see Lawrence et al., 2007; Strutt, 2008; Axelrod, 2009; Bayly and Axelrod, 2011), and as we discuss later, it remains unclear whether Ft/Ds signalling feeds directly or indirectly into core protein localisation, or if it in fact represents an independent and parallel system for setting up planar polarity. Whether Ft/Ds signalling does act as a long-range polarity signal, and the mechanisms by which such a signal might be established and transduced are also important questions that we examine in this review.

THE FT/DS SYSTEM AND PLANAR POLARITY

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. THE FT/DS SYSTEM AND PLANAR POLARITY
  5. THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?
  6. DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION
  7. INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING
  8. ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?
  9. PERSPECTIVES
  10. Acknowledgements
  11. REFERENCES

The ft, ds and fj genes were identified many years ago on the basis of defects in growth and morphogenesis of Drosophila tissues in mutants (Mohr, 1923, 1929; Stern and Bridges, 1926; Waddington, 1943; Bryant et al., 1988), and were later found to be also involved in planar polarity (Adler et al., 1998; Zeidler et al., 1999). Acting together as one of the upstream inputs into the Hpo/Wts tumour suppressor pathway, they prevent Yorkie (Yki)-mediated transcription of genes involved in cell cycle progression, and hence restrict tissue size (Fig. 3A, reviewed in Grusche et al., 2010; Oh and Irvine, 2010). The overgrowth of appendages seen in mutants is not uniform, but is particularly marked perpendicular to the proximo-distal axis indicating that Ft/Ds signalling normally precisely coordinates growth and morphogenesis to produce tissues of the correct size and shape (Fig. 3D,E, Bryant et al., 1988; Clark et al., 1995; Garoia et al., 2000; Baena-López et al., 2005).

thumbnail image

Figure 3. Functions of Ft/Ds signalling. A: The Ft/Ds signalling pathway in growth. The activity and properties of Ft are modified by the action of Ds, Fj, Lowfat (Lft), and Discs overgrown (Dco). Ft suppresses D activity, which represses Wts/Hpo signalling, which in turn prevents Yki from entering the nucleus and promoting transcription of genes involved in cell cycle progression and growth. Approximated (App) is involved in promoting the localisation of D (see text for further details). Various other inputs (not discussed in this review) feed into the pathway. B: A lateral view of the interaction between Ft (cyan) and Ds (magenta). Ft and Ds localise apically in cells where they bind heterophilically through their extracellular cadherin domains. The affinity of binding is modulated by the action of Fj in the Golgi (yellow circles), which promotes Ft binding to Ds, but inhibits Ds binding to Ft. C: Localisation of Ft, Ds, and D in the wing. This cartoon represents an apical view, with proximal to the left. Ft and Ds (cyan and magenta hatching) are localised at the membrane with no apparent polarity. D (green) is localised to the distal side of cells. D–F: Growth phenotypes seen in Drosophila wings mutant for components of the Ft/Ds pathway. Wings are oriented with proximal to the left and anterior upwards. Compared to the wild-type (D), wings are enlarged in the antero-posterior axis in ds mutants (E) and reduced in size in d mutants, particularly along the proximo-distal axis (F). Images (D–F) courtesy of A. Brittle.

Download figure to PowerPoint

Ft and Ds are both large atypical cadherins that localise just apical to epithelial adherens junctions and bind to one another heterophilically across cell boundaries (Fig. 3B, Mahoney et al., 1991; Clark et al., 1995; Strutt and Strutt, 2002; Yang et al., 2002; Ma et al., 2003; Matakatsu and Blair, 2004). Their activity and accumulation is influenced both by their interactions with each other and through the action of several other proteins (Fig. 3A). The most important of these (at least as far as planar polarity is concerned) is Fj, which is a Golgi kinase that phosphorylates the extracellular cadherin domains of both Ft and Ds, thus modifying their heterophilic binding such that the ability of Ft to bind Ds is increased, but the ability of Ds to bind Ft is reduced (Yang et al., 2002; Strutt et al., 2004; Ishikawa et al., 2008; Brittle et al., 2010; Simon et al., 2010). However, Ft and Ds levels are also modified by Lowfat (Lft), a poorly understood homologue of human LIX1, which binds to the intracellular domains of both cadherins and stabilises them (Mao et al., 2009). Ft is also constitutively cleaved, and its cytoplasmic domain is phosphorylated in a Ds-dependent manner by the action of the kinase Discs overgrown (Dco) (Feng and Irvine, 2009; Sopko et al., 2009), although the significance of the former modification is unknown and the latter seems to predominantly affect the influence of Ft in growth rather than polarity.

The defects observed in mutants indicate that the Ft/Ds system has a non-autonomous effect on planar polarity, separable from but potentially acting upstream of core protein activity in some tissues (Adler et al., 1998; Rawls et al., 2002; Strutt and Strutt, 2002; Yang et al., 2002; Ma et al., 2003). Analysis in the Drosophila eye has established that mutation of core planar polarity proteins such as Fz, disrupts asymmetric localisation of other core proteins causing defects in direction, degree, and chirality of rotation (Fig. 1G, Zheng et al., 1995; Wolff and Rubin, 1998; Das et al., 2002; Strutt et al., 2002; Yang et al., 2002), whereas tissue mutant for ft, ds, or fj retains core protein asymmetry and specifically results in chirality inversions (Fig. 1F,H; Zeidler et al., 1999; Yang et al., 2002). Both fz and genes of the Ft/Ds system show cell non-autonomous phenotypes, as chirality defects in mutant clones can be propagated into wild-type tissue on one side of the clone (Fig. 1G,H; Zheng et al., 1995; Zeidler et al., 1999; Rawls et al., 2002; Strutt et al., 2002; Yang et al., 2002). However, ft mutant clones are unable to influence chirality in tissue mutant for core polarity genes indicating that the Ft/Ds system appears to be acting epistatically upstream of the core system in the eye (Yang et al., 2002). Ft/Ds-mediated directional non-autonomy can also be seen in the wing and in other Drosophila tissues such as the adult abdomen and the larval epidermis where planar polarity is required for correct orientation of bristles and denticles (Adler et al., 1998; Zeidler et al., 2000; Casal et al., 2002; Strutt and Strutt, 2002; Ma et al., 2003; Lawrence et al., 2004; Repiso et al., 2010; Donoughe and DiNardo, 2011). Notably, although ft is expressed uniformly along the dorso-ventral axis of the eye primordium, ds and fj are expressed in gradients, peaking respectively in polar and equatorial regions due to regulation by Wingless (Wg) signalling at the poles (see Fig. 6A; Clark et al., 1995; Villano and Katz, 1995; Brodsky and Steller, 1996; Zeidler et al., 1999; Yang et al., 2002). Given that classical models of planar polarity had long considered the idea of gradients to provide a potential explanation for the establishment of long-range polarity (Locke, 1959; Lawrence, 1966; Stumpf, 1966; Lawrence et al., 1972), these data led to the proposal that the Ft/Ds system might act to globally coordinate core protein-mediated polarity with the body axes (Casal et al., 2002; Yang et al., 2002; Ma et al., 2003).

THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. THE FT/DS SYSTEM AND PLANAR POLARITY
  5. THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?
  6. DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION
  7. INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING
  8. ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?
  9. PERSPECTIVES
  10. Acknowledgements
  11. REFERENCES

The model proposed by Yang et al. (2002) suggested that in the eye, a gradient of Ft function set up by the action of graded ds and fj expression might be directly interpreted by Fz-mediated comparison of Ft activity levels between the presumptive R3 and R4 photoreceptors. Since then, in the eye itself there has been little evidence to contradict this hypothesis, but equally no molecular link between the two systems has been found and an increasing number of results in other Drosophila tissues point toward the Ft/Ds system having an indirect effect on core protein-mediated polarity. In common with the eye, other tissues studied show a relationship between the direction of Ft/Ds polarity and core protein polarity (Adler et al., 1998; Strutt and Strutt, 2002; Ma et al., 2003; Casal et al., 2006). However, intriguingly there seem to be differences in interpretation of the Ft/Ds polarity signal according to the specific context, as in some cases (such as the eye and the anterior segment compartments of the abdomen) core planar polarity points towards high ds expression (Yang et al., 2002; Casal et al., 2006), and in others (such as the wing and posterior segment compartments in the abdomen) towards low ds expression (Adler et al., 1998; Strutt and Strutt, 2002; Ma et al., 2003; Casal et al., 2006), meaning a direct relationship between Ft/Ds activity levels and the asymmetry of core proteins is unlikely.

The complexity of the interaction between the two systems is illustrated in the Drosophila wing where core protein-mediated planar polarity is required for the orientation of both trichomes and cuticle ridges (Wong and Adler, 1993; Doyle et al., 2008). Although core protein localisation is fundamentally involved in the mechanics of trichome placement (Adler et al., 2004; Strutt and Warrington, 2008), indicating that at some point Ft/Ds signalling must be acting upstream of the core in trichome orientation, there are several pieces of evidence that indicate that the interaction is indirect. In particular, trichome polarity defects are only seen in ft or ds mutant clones that are particularly large or proximal, and core proteins are not dramatically mislocalised around clones as would be expected if the interaction was direct (Strutt and Strutt, 2002; Ma et al., 2003). Intriguingly, core protein activity has been shown to operate in two, temporally separable phases during pupal development, one occurring sometime before 18h APF, the other between 18–32h APF, each having different requirements in its use of the cytoplasmic components of the core polarity system (Strutt and Strutt, 2002, 2007; Doyle et al., 2008). Early core protein activity is required to align polarity with the body axes, and eliminating this early function, then restoring it after 18h APF results in global polarity defects that strongly resemble those seen in ft or ds mutant wings (Strutt and Strutt, 2002, 2007), suggesting the two pathways might be connected at this stage. Further evidence for this comes from analysis of cuticle ridges, as the orientation of posterior ridges is set up by the early phase of core protein activity, whereas the later phase orients the anterior ridges such that they run perpendicular to the posterior ridges (Doyle et al., 2008). Ft/Ds signalling acts only in posterior ridge orientation indicating that there is both spatial and temporal control of the interaction between the two polarity systems (Hogan et al., 2011).

A linear pathway between Ft/Ds signalling and the core proteins also seems unlikely to exist in either the abdomen or the larval epidermis, where instead the two systems appear to act cooperatively to set up polarity in a parallel fashion (Casal et al., 2006; Repiso et al., 2010; Donoughe and DiNardo, 2011). In the abdomen, gradients of Hedgehog (Hh) and Wg morphogens arising, respectively, from the posterior and anterior compartments, regulate both Ft/Ds and core protein activity (Struhl et al., 1997; Lawrence et al., 1999; Casal et al., 2002, 2006; Lawrence et al., 2002), and planar polarity must be coordinated through integration of the two polarity systems, although how this is mediated is unknown. A similar system appears to operate in the larval epidermis, but additionally there appears to be a high level of redundancy as mutating the core system alone has little effect (Repiso et al., 2010). However, the enhanced polarity defects seen in ds; fz double mutant larvae compared with ds single mutants indicates that both systems are acting in parallel to pattern spatially distinct domains of each embryonic compartment (Donoughe and DiNardo, 2011).

These conflicting results suggest that the Ft/Ds system has different roles in different tissues, and that signalling may be directly or indirectly upstream of the core system, independent or cooperative depending upon tissue requirements. Most importantly, the results in the abdomen and larval epidermis indicate that the Ft/Ds system represents a viable polarity-generating system in its own right, and hence it may be best to think of Ft/Ds and the core as two separate means of generating polarity, that come together in different ways under different circumstances. Of course, this complicates the process of determining the mechanism of any interaction between the two systems, as not only might there be tissue specificity, but also the lack of consistent linearity in the pathway indicates that there are potentially a lot more gaps in our knowledge than previously thought. We now need to search not only for mechanisms that mediate interactions between Ft/Ds and the core, but also the means of setting up global core planar polarity in a Ft/Ds-independent manner, and setting up planar polarity downstream of Ft/Ds signalling in a core protein-independent manner.

DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. THE FT/DS SYSTEM AND PLANAR POLARITY
  5. THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?
  6. DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION
  7. INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING
  8. ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?
  9. PERSPECTIVES
  10. Acknowledgements
  11. REFERENCES

Some progress towards determining how a Ft/Ds-dependent polarity signal might be transduced has come from analysis of the atypical myosin Dachs (D), a downstream effector of Ft/Ds signalling in growth regulation that binds to and inhibits the activity of Wts (Cho and Irvine, 2004; Cho et al., 2006; Mao et al., 2006; Milton et al., 2010). Intriguingly, D itself is localised asymmetrically at the presumptive distal side of cells in wild-type wing discs (the larval precursors of the adult wing) in a Ft/Ds-dependent manner (Fig. 3C, Mao et al., 2006; Rogulja et al., 2008). Data from clonal analysis in the wing show that cells ectopically expressing Ds or lacking Ft activity, recruit Ft to the cell membrane in surrounding wild-type cells through heterophilic adhesion (Strutt and Strutt, 2002; Ma et al., 2003; Matakatsu and Blair, 2004) and within such clones D is strongly recruited asymmetrically to clone boundaries where Ds is localised (Mao et al., 2006). Whether D is recruited directly through Ds or indirectly by Ft/Ds activity is unclear, but its localisation is dependent upon the activity of Approximated (App), a DHHC (Asp-His-His-Cys) palmitoyltransferase (Matakatsu and Blair, 2008). Although palmitoylation is known to regulate membrane targeting and intracellular trafficking of proteins (reviewed in Greaves et al., 2009; Salaun et al., 2010), D itself does not appear to be the target of App palmitoylation (Matakatsu and Blair, 2008), and it is not known whether App acts to mediate the effects of Ft/Ds signalling on D, or whether it represents an independent way to regulate D accumulation and activity.

D asymmetry can be thought of as a form of coordinated cell-cell planar polarity established directly downstream of Ft/Ds signalling. Interestingly, the asymmetric localisation of D appears to be specifically required to regulate planar polarised tissue growth. Evidence from making GFP-marked clones in wild-type wings indicates that the direction of growth is biased distally, predominantly due to cell divisions being preferentially oriented along the proximo-distal axis (Baena-López et al., 2005). However, ft mutant clones are both larger than wild-type and rounded instead of elongated (Baena-López et al., 2005; Mao et al., 2006, 2011b), suggesting that cell division is not only excessive but also mispolarised in the absence of long-range polarity information normally supplied by gradients or boundaries of Ft activity. Recent work has shown that D is able to mediate both of these effects, the first transcriptionally through Yki de-repression (Cho et al., 2006), the second through alignment of the mitotic spindle, which correlates with D asymmetric localisation and orientation of cell division (Fig. 4, Mao et al., 2011b). Spindles are known to form preferentially along the longest planar axis of epithelial cells prior to division (Lu et al., 2001; Théry et al., 2007; Vogel et al., 2007), and it appears that the myosin activity of D may promote cell elongation by distally constricting apical junctions of cells in which it is asymmetrically localised (Mao et al., 2011b).

thumbnail image

Figure 4. Dachs promotes polarised cell division. A: In cells without asymmetric D, the mitotic spindle (red) aligns randomly and cell division can occur in any direction, resulting in clones with a rounded shape. B: Asymmetric D (green) constricts the cell membrane, meaning cells become elongated. Spindles preferentially align along the long axis of cells, thus resulting in polarised division and elongation of clones and tissues.

Download figure to PowerPoint

Although the polarity of D and of core proteins is often correlated (for example, both D and Fz are localised towards the presumptive wing margin in larval and early pupal wing cells, A. Brittle, C. Thomas and D. Strutt, unpublished data, Classen et al., 2005; Mao et al., 2006; Rogulja et al., 2008; Aigouy et al., 2010; Strutt et al., 2011), d mutants exhibit little in the way of classical, core-dependent planar polarity defects (Mao et al., 2006), indicating that D asymmetry is likely only required to transmit a subset of Ft/Ds mediated polarity signals and is not required for polarisation of the core system. However, as d mutant wings also undergo fewer cell divisions and are reduced in size (Fig. 3F, Mao et al., 2006), it is possible that in wild-type wings, those cell divisions that are mediated by D activity must be carefully polarised to prevent disorganisation (both of shape and polarity) that would otherwise be created if uncoordinated cell proliferation were to occur. In other words, D may be required for maintenance of core protein-mediated planar polarity as the wing grows in a D-dependent manner, but is probably not involved in the processes that lead to initiation of core asymmetry. In ft mutant tissue, D is active but not asymmetric, and there are both excess, unpolarised cell divisions and severe core protein-dependent polarity defects (Bryant et al., 1988; Strutt and Strutt, 2002; Yang et al., 2002; Baena-López et al., 2005; Mao et al., 2011b). However, whether the polarity defects are caused primarily by the unpolarised cell divisions, or whether Ft has other, D-independent means of establishing core polarity is still unclear.

Another molecule that putatively acts downstream of Ft/Ds signalling is the Drosophila homologue of Atrophin (Atro), known also as Grunge (Gug), which has been found to interact both physically and genetically with Ft (Fanto et al., 2003). Unlike D, Atro does not appear to be involved in Ft-mediated growth regulation, but loss-of-function mutant tissues do show cell non-autonomous polarity defects that resemble the loss of ft (Zhang et al., 2002; Fanto et al., 2003), and there is some evidence that like Ft, Ds, and D, Atro is required to orient cell divisions in the Drosophila wing (Li et al., 2009). Atro functions as a transcriptional co-repressor (Erkner et al., 2002; Zhang et al., 2002), and its effects on planar polarity may be explained at least in part by its transcriptional control of polarity genes such as fj (Fanto et al., 2003). However, as Atro also regulates the expression of many other developmental genes, the analysis of polarity phenotypes has been hampered by pleiotropic effects, meaning that its relationship to other Ft/Ds pathway members and the significance of its binding with Ft, are still poorly understood.

INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. THE FT/DS SYSTEM AND PLANAR POLARITY
  5. THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?
  6. DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION
  7. INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING
  8. ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?
  9. PERSPECTIVES
  10. Acknowledgements
  11. REFERENCES

Recent work in the wing has suggested several alternate mechanisms by which Ft/Ds signalling might indirectly affect core protein-mediated trichome polarity, although whether these operate through D or alternative factors is unknown at this point. Proximo-distal asymmetry of core planar polarity proteins peaks at 30h APF just before trichomes form (Usui et al., 1999; Axelrod, 2001; Strutt, 2001; Classen et al., 2005; Aigouy et al., 2010), and this correlates with the requirement of core signalling activity to generate the strong subcellular asymmetry necessary to direct polarised hair formation (Strutt and Strutt, 2002). However, core planar polarity proteins also exhibit some asymmetry in larval and early pupal wings, which is then disrupted due to cell rearrangements between 15h and 24h APF (Classen et al., 2005; Aigouy et al., 2010; Strutt et al., 2011), and as we previously mentioned, there is a requirement for core protein signalling prior to 18h APF (Strutt and Strutt, 2002; Strutt and Strutt, 2007; Doyle et al., 2008). Unlike the later, distally oriented asymmetry, early polarity of core proteins appears to point either anteriorly or posteriorly towards the wing margins, radiating out from the third wing vein (L3) (Fig. 5A; Classen et al., 2005; Aigouy et al., 2010; Strutt et al., 2011). In fact, it has been observed that Ds is briefly expressed along L3 at 17h APF and some results indicate that it might act as a polarising signal at this time both for trichome polarity and wing ridge alignment (Adler et al., 1997; Matakatsu and Blair, 2004; Hogan et al., 2011). However, given that core proteins are polarised in the same manner in 6h APF pupal wings (Strutt et al., 2011), and show some asymmetry even earlier than this in larval wing discs (Classen et al., 2005), it is probable that the overall pattern of core planar polarity is set up much earlier, but perhaps maintained or influenced by other signals later in development.

thumbnail image

Figure 5. Functions of Ds in planar polarity in the wing. In all panels proximal is to the left, anterior (A) is upwards, and posterior (P) is downwards. A: At 15h APF, the wing is planar polarised (red arrows) away from the third wing vein (L3, blue dashed line). Hinge morphogenesis creates tension in the wing, resulting in a flow of cells (wide grey lines with arrowheads) towards the hinge at L3, and away from the hinge in posterior and anterior sections of the wing. Polarity becomes reoriented to point distally. This movement is partially regulated by Ds, which promotes cell elongation and polarised cell division. B: A cartoon showing a single cell in the anterior of the wing as it undergoes the process described in A. At 15h APF, microtubules are aligned generally along the antero-posterior axis, but are not oriented uniformly and there is no bias in the direction of Fz transport (green circles). Fz (green bar) shows asymmetry towards the anterior. As wing morphogenesis takes place, the cell rotates clockwise (wide grey line with arrowheads). Microtubule alignment and orientation become consolidated in a Ds-dependent manner, so that by 30h APF, microtubules are preferentially aligned along the proximo-distal axis, with a predominance of plus ends facing towards the distal side of the cell. Fz is transported towards plus ends where it accumulates distally, resulting in a distally positioned trichome.

Download figure to PowerPoint

How does anterior/posterior core planar polarity in early pupal wings translate into later distal polarity? Work from Suzanne Eaton's group has shown that between 15h and 24h APF, there is a proximal flow of wing cells centred on L3, which acts both to elongate the wing and rotate the axis of polarity (Fig. 5A, Aigouy et al., 2010). This is driven by contraction of the hinge, which increases tension along proximo-distal wing cell boundaries, initially resulting in cell elongation and oriented cell division and later promoting rearrangement of cells to alleviate energetically unstable cell geometries. Although the signals behind hinge contraction are unknown, Ds plays a key role in this process by promoting cell elongation and oriented cell division (Aigouy et al., 2010), presumably at least partially through regulation of D asymmetry, although this has not been investigated. Not only does this mechanism act to reorient polarity distally, it also has a tendency to create new antero-posterior boundaries, thus promoting accumulation of core proteins on the pre-existing proximo-distal cell boundaries (Aigouy et al., 2010).

Another mechanism for consolidation of core protein-mediated polarity that seems to be dependent upon Ft/Ds signalling is the polarisation of non-centrosomal microtubules during pupal wing development (Fig. 5B). In 30h APF pupal wing cells, microtubules are organised roughly in proximo-distal arrays localised at the level of apical adherens junctions, and a slight distal bias of microtubule plus ends has been proposed to influence the transport of Fz sufficiently to promote its accumulation at the distal end of the cell (Shimada et al., 2006). Microtubules appear to reorient in a Ft/Ds dependent manner, from a predominantly antero-posterior direction at 14h APF, to a more proximo-distal orientation by 24h APF (Harumoto et al., 2010). It might be expected in light of the data discussed above, that the reorientation is simply due to the Ds-dependent cell rearrangement that is occurring at this time. However, in ds mutant wings, early antero-posterior orientation is also disrupted, and the later distal bias of plus ends is not seen, suggesting that Ft/Ds signalling also has a more direct effect on microtubule polarisation, perhaps through influencing the subcellular localisation of Par-1 (Harumoto et al., 2010). The role of D in this process has not been investigated but it is tempting to speculate that it could be involved in orienting microtubules, either through tethering of plus ends, or more indirectly through its control of cell elongation.

So far, all of these data are pointing towards a complex, indirect effect of Ft/Ds signalling in the wing, where control of cell elongation, oriented cell division, and perhaps other unknown processes has a wide-ranging series of effects on wing growth, epithelial morphogenesis, and cytoskeletal architecture, which come together to influence, coordinate, and consolidate core protein-mediated polarity. Thus these findings show how polarity is maintained and strengthened whilst the wing is undergoing various morphogenetic processes, but still leave open the question of how core planar polarity is initially set up in this tissue. In particular, it is still unclear whether Ft/Ds signalling could set up core polarity via a D-independent mechanism, or whether core polarity might be initiated by an additional, as yet uncharacterised global signal.

ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. THE FT/DS SYSTEM AND PLANAR POLARITY
  5. THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?
  6. DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION
  7. INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING
  8. ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?
  9. PERSPECTIVES
  10. Acknowledgements
  11. REFERENCES

The molecular properties of Ft, Ds, and Fj provide a basis by which signals can be modified and transmitted from one cell to the next, and together with genetic data have led to the development of two alternative models for polarity propagation: the “gradient model” and the “boundary-relay model,” both of which additionally explain how Ft/Ds activity can be coupled to the expression patterns of the morphogens that pattern the developing tissues.

The Gradient Model

As discussed briefly earlier in the review, the idea of a graded signal as providing a mechanism for long-range patterning and polarity had been discussed and developed for many years (Locke, 1959; Lawrence, 1966; Stumpf, 1966; Lawrence et al., 1972). Therefore, it was a natural step following the discovery of the graded expression patterns of ds and fj in the Drosophila eye and abdomen (Fig. 6A,B) and the non-autonomous phenotypes seen in mutant clones (Fig. 1H; Villano and Katz, 1995; Brodsky and Steller, 1996; Zeidler et al., 1999; Casal et al., 2002; Yang et al., 2002; Lawrence et al., 2004), to propose a gradient model for Ft/Ds-mediated planar polarity propagation. Although eliminating either the Ds or Fj gradients has little effect on polarity in the eye, levelling both gradients (and therefore levelling a presumptive Ft activity gradient also) by uniformly expressing both genes in genetic null backgrounds, results in a complete disruption of planar polarity (Simon, 2004). In the abdomen, a gradient model for long-range polarity has also been favoured. Here ds and fj expression gradients are segmentally repeated in response to Wg and Hh morphogen gradients (Fig. 6B), and as in the eye, high Wg signalling corresponds to high ds and low fj expression (Struhl et al., 1997; Casal et al., 2002; Lawrence et al., 2002).

thumbnail image

Figure 6. Gradient and boundary-relay models for propagation of long-range polarity. Gradients of ds (magenta) and fj (yellow) have been postulated to mediate long-range polarity in the Drosophila eye disc (A) and adult abdomen (B), whilst in the wing disc (C) a boundary-relay mechanism seems more likely. Note that the direction of core planar polarity (navy arrow) is not consistent with the expected slope/boundary of ds and fj expression, suggesting an indirect input of the Ft/Ds system. A, anterior; P, posterior; D, dorsal; V, ventral. Di: A representation of the gradient model for propagation of long-range polarity. ds expression is higher to the left (which could represent proximal in the wing or polar in the eye for example), whilst fj expression is higher to the right. ft expression is uniform, but the ability of Ft to bind Ds is promoted in a graded manner (from pale to dark cyan) by Fj activity. Each cell reads its position in the gradient through Ft/Ds heterophilic binding. D (green) becomes localised to the side of each cell with higher Ds/lower Ft activity. Dii: A clone (second cell from left) of ds mutant cells within the gradient causes a unidirectional polarity reversal. Ft in wild-type cells abutting the clone cannot interact heterophilically with ds mutant cells, so redistributes on the side of the cell away from the clone boundary. The polarity reversal (represented by D localisation) is propagated for several cell diameters, but is gradually diluted through the influence of the endogenous gradient. E: A representation of the boundary-relay model for propagation of long-range polarity in the wing. Hinge cells expressing high ds are unpolarised and Ft in these cells binds Ds poorly due to lack of Fj activity. fj expression is upregulated in the wing primordium through the action of vg, and promotes the ability of Ft to bind Ds. Strong heterophilic interactions between Ds and Ft occur at the hinge/wing boundary (black dashed line), causing Ds and Ft to become polarised in cells flanking the boundary and D to be recruited asymmetrically. Signalling through D promotes transcription of vg, which in turn activates fj and represses ds expression, thus moving the expression boundary forward by one cell diameter (solid black arrow). The process is then repeated, resulting in polarisation of the wing blade as it is recruited. As the expression boundary moves away, cells gradually lose their strong polarisation of Ft and Ds, but D asymmetry is maintained.

Download figure to PowerPoint

How would a gradient system work at the molecular level? In the model proposed by Lawrence and colleagues (2004), individual cells determine their position in the gradient according to absolute levels of activated Ft, but can also read the position of neighbouring cells and hence the slope of the gradient through Ft/Ds heterophilic interaction (Fig. 6Di). This would result in subtle asymmetries in the number or activity of heterodimers on each side of the cell. Discontinuities in the gradient, for example ds loss-of-function clones (Fig. 6Dii), would be sensed by adjacent wild-type cells that would then no longer be able to form heterophilic interactions with cells inside the clone, and so would redistribute Ft towards the boundary with other wild-type cells where Ds is present. This asymmetry of protein localisation would effectively reverse the gradient at one side of the clone, resulting in a polarity inversion that would then propagate for several cells. Indeed, redistribution of Ft and Ds proteins on the edges of loss-of-function/over-expression clones has been noted by several groups, and the corresponding alterations in polarity correlate with that expected from the gradient model (Ma et al., 2003; Mao et al., 2006).

The gradient model represents a simple method for translating long-range morphogenetic signals (e.g., Hh and Wg) into global polarity signals, and can satisfactorily explain experimental data obtained from clones. However, as gradients are generally shallow in wild-type tissue, such a system can only serve to provide a subtle polarity cue to any individual cell, and consistent with this neither Ft nor Ds shows obvious asymmetry of localisation. It is possible that cells could read the gradient early in development when the tissue is small and gradients are steeper, and then in some way retain the polarity information as the tissue grows. As Ft/Ds signalling is also required to regulate tissue growth, this would allow simultaneous coordination of both processes (discussed further in Lawrence et al., 2008). Additionally, there must be some type of downstream feedback system that is able to accurately amplify small disparities in Ft/Ds protein localisation to generate a more robust cellular polarity cue such as the asymmetry of D or of core proteins.

The gradient model encounters some problems in the Drosophila wing, however, where there are several pieces of evidence indicating that Ft/Ds activity gradients have little influence in determining planar polarity, even though the generation of ectopic gradients is sufficient to alter the direction of polarity (Matakatsu and Blair, 2004). In particular, despite ds and fj expression patterns being mutually exclusive in the wing (Clark et al., 1995; Zeidler et al., 2000), the distribution of Ds is not obviously graded throughout the wing blade (Strutt and Strutt, 2002). Furthermore, gradients of these molecules do not appear to be required for core planar polarity in most of the wing as uniform expression of ds and fj in ds fj double mutant wings results in rescue of all but the most proximal polarity defects (Matakatsu and Blair, 2004; Simon, 2004; Brittle et al., 2010). One point to note is that although ft expression appears to be fairly uniform in the wing (Garoia et al., 2000), Ft protein levels are augmented at the dorso-ventral boundary of the larval wing pouch due to the stabilising action of Lft (Strutt and Strutt, 2002; Mao et al., 2009), thus potentially providing some polarity information in the absence of ds and fj gradients. Additionally, the presence in the wing of as yet unidentified graded factors that modulate Ft function cannot be ruled out.

The Boundary-Relay Model

An alternative to the gradient model for the long-range propagation of polarity, which may be more relevant to the wing, has been developed based upon consideration of expression boundaries rather than gradients. As mentioned previously, experimentally induced sharp discontinuities in Ft/Ds signalling (as seen at the edge of loss-of-function or over-expression clones) cause D to be recruited asymmetrically (Mao et al., 2006) and to provoke non-autonomous polarity defects that can propagate for several cell diameters (Fig. 1H, Adler et al., 1998; Zeidler et al., 1999, 2000; Casal et al., 2002; Strutt and Strutt, 2002; Yang et al., 2002; Ma et al., 2003; Repiso et al., 2010; Donoughe and DiNardo, 2011). Since growth through the Hpo/Wts pathway is promoted in the cells at such boundaries (Rogulja et al., 2008; Willecke et al., 2008), it has been suggested that in wild-type animals this may represent a method of mediating regenerative tissue growth after damage or wounding (Grusche et al., 2011). However, expression boundaries are also seen naturally in tissues during development. In the wing disc, for example, the wing selector gene vestigial (vg) activates fj in the wing pouch and represses ds (Zecca and Struhl, 2010; Schwank et al., 2011), resulting in its restriction to the hinge region and creating a boundary between high ds and high fj expression (Fig. 6C).

Static expression boundaries may be sufficient to direct polarity in adjacent cells, but such a mechanism is unlikely to be able to mediate global polarisation of any but the smallest tissues. In the wing, this problem is circumvented by the presence of an Ft/Ds-dependent relay, termed the “feed forward” mechanism (Fig. 6E, Zecca and Struhl, 2010). Initially, the presumptive wing comprises only a few cells expressing vg at the dorso-ventral midline in response to Notch signalling (Kim et al., 1996). However, wing cells are progressively recruited through an auto-regulatory loop of vg induction, in response to Wg long-range and Ft/Ds short-range signalling (Zecca and Struhl, 2007a, b, 2010). In receiving cells, this process is dependent upon D (Zecca and Struhl, 2010), which is likely to become asymmetrically localised with each round of cell recruitment as a new Ft/Ds activity boundary is established. Indeed, D asymmetry has been observed in third instar wing discs (Mao et al., 2006; Rogulja et al., 2008) and correlates with the pattern expected by the feed-forward process, indicating that the asymmetry is maintained as the tissue grows and develops.

Could this boundary-relay mechanism be a means of setting up long-range polarity as well as expanding the wing primordium? Further investigation is needed, but in terms of D asymmetry, which is currently the only known manifestation of polarity that is directly downstream of Ft/Ds signalling, this is an attractive model (Fig. 6E). In particular, it suggests that polarity could be established right from the beginning of wing development, then maintained until adulthood, rather than being generated de novo at some unknown point in early larval life. One point to note is that even though D is required to transduce the feed-forward mechanism (Zecca and Struhl, 2010), d mutant wings show very few polarity defects and although small, do grow to a certain extent (Fig. 3F; Mao et al., 2006). This indicates that feed-forward alone probably only accounts for a proportion of wing recruitment and growth, and similarly may only be involved in setting up polarity in several cell rows, although this could then be propagated further throughout the tissue. Supporting this prediction, D asymmetry in wing discs is more marked in presumptive proximal regions (A. Brittle, C. Thomas, and D. Strutt, unpublished observations), and it is the proximal part of the wing that remains susceptible to polarity defects when the fj and ds expression patterns are levelled (Matakatsu and Blair, 2004; Simon, 2004; Brittle et al., 2010).

Although a boundary mechanism obviates the necessity for the initial amplification of asymmetry required in a gradient model, instead a memory of the original cell polarity must be maintained throughout development. This would likely require a similar sort of feedback system to that necessary for amplification of asymmetry, meaning that both models depend upon additional factors, as yet undiscovered. Furthermore, in tissues such as the eye and abdomen, clear Ft/Ds signalling boundaries have not been reported, although this may be partly because gradient (rather than boundary) models have been favoured by those working on such tissues. However, in both tissues it is evident from clonal data that induction of ectopic boundaries is sufficient to alter long-range polarity (Casal et al., 2002; Yang et al., 2002), and in fact the distinction between gradients and boundaries is not always clear, given that a boundary between fj and ds expressing cells can be considered an extreme form of opposing expression gradients. Thus it is possible that both gradient and boundary signals use similar cellular machinery to set up global polarity in different circumstances, depending upon the morphogenetic history of the tissue being patterned.

PERSPECTIVES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. THE FT/DS SYSTEM AND PLANAR POLARITY
  5. THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?
  6. DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION
  7. INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING
  8. ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?
  9. PERSPECTIVES
  10. Acknowledgements
  11. REFERENCES

There are clearly many questions that remain to be answered in determining how Ft/Ds signalling sets up long-range polarity, the actual nature of the cellular polarity cues generated, and how these impact upon core protein-mediated polarity.

The Drosophila wing was initially chosen for polarity analysis due to its deceptively simple two-dimensional array of near identical epithelial cells. However, we now know that in fact this regular epithelial sheet is formed through complex processes of growth, and cell flow and rearrangement, all of which are strongly influenced by Ft/Ds signalling and difficult to separate out from other effects that the Ft/Ds system might be having on polarity. It may be that another tissue such as the larval epidermis where polarity is generated with reference to only a single (antero-posterior) axis, is a simpler model system to use, although as the two systems seem to act independently in this tissue it seems unlikely that this is going to tell us much about how Ft/Ds signalling impacts on the core. Alternatively, a better model system may be the early pupal wing, or even the wing disc, before the complex morphogenetic processes of disc inversion and wing elongation take place.

In terms of generating long-range polarity, the boundary-relay mechanism needs to be investigated further. In particular, it is important to ascertain whether it is a viable mechanism in tissues other than the wing, through investigation of potential boundaries such as the antero-posterior border in the adult abdomen and larval epidermis, perhaps with the help of D localisation as a marker for Ft/Ds-mediated asymmetry. Investigating the mechanisms by which D becomes asymmetric, whether by direct physical interaction with Ft or Ds, or indirect recruitment or amplification, will also be useful, not only to discover how asymmetry is generated, but also to uncover novel outputs of Ft/Ds-generated polarity that will help us to understand in more depth how the signal is transmitted and interpreted.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. THE FT/DS SYSTEM AND PLANAR POLARITY
  5. THE FT/DS AND CORE PLANAR POLARITY SYSTEMS: LINEAR OR PARALLEL POLARITY PATHWAYS?
  6. DOWNSTREAM OF FT/DS SIGNALLING: DACHS, ATROPHIN AND POLARISED CELL DIVISION
  7. INDIRECT EFFECTS OF THE FT/DS SYSTEM ON CORE POLARITY IN THE WING
  8. ESTABLISHING FT/DS-MEDIATED GLOBAL POLARITY: GRADIENTS OR BOUNDARY-RELAY MECHANISMS?
  9. PERSPECTIVES
  10. Acknowledgements
  11. REFERENCES