Receptor Tyrosine Kinases in Drosophila Development

Abstract

Tyrosine phosphorylation plays a significant role in a wide range of cellular processes. The Drosophila genome encodes more than 20 receptor tyrosine kinases and extensive studies in the past 20 years have illustrated their diverse roles and complex signaling mechanisms. Although some receptor tyrosine kinases have highly specific functions, others strikingly are used in rather ubiquitous manners. Receptor tyrosine kinases regulate a broad expanse of processes, ranging from cell survival and proliferation to differentiation and patterning. Remarkably, different receptor tyrosine kinases share many of the same effectors and their hierarchical organization is retained in disparate biological contexts. In this comprehensive review, we summarize what is known regarding each receptor tyrosine kinase during Drosophila development. Astonishingly, very little is known for approximately half of all Drosophila receptor tyrosine kinases.

About 20 RTKs are encoded by the Drosophila genome, most of which have a mammalian counterpart. They provide a means of communication between different tissues and cell types that leads to a robust and reproducible developmental program.

One of the key strategies that arose during evolution to facilitate the transmission of extracellular information was that of receptor tyrosine kinase (RTK) signaling. This mechanism enables cells to transduce cues from their extracellular environment and thus contributes extensively to developmental processes. Today, we have come to recognize conserved RTK signaling as crucial for most aspects of cell fate determination, differentiation, patterning, proliferation, growth, and survival in metazoans. Activation of RTKs by ligand leads to a canonical deployment of signal transduction involving adaptor proteins, serine/threonine kinases, and transcription factors essential for animal development.

RTKs function reiteratively in different contexts during development to direct, restrain, or alter the commitment of a cell. Genetically tractable model organisms such as Drosophila melanogaster have proven instrumental in deciphering the roles of RTKs during development as well as their signaling pathways. Furthermore, extension of this knowledge to mammalian orthologs has substantially broadened our understanding of the function of RTKs in development and cellular transformation. Approximately 20 RTKs are encoded by the Drosophila genome, nearly all of which have a mammalian counterpart (Table 1). In this article, we review what we know to date about their functions, illustrating the diversity of cellular processes controlled by RTK signaling as well as the extent of pleiotropy associated with specific RTKs.

Table 1.

Drosophila RTKs/ligands/signaling components/transcription factors

Flybase ID	Symbol	Name	Mammalian homolog	Ligand	Characterized signaling pathway components
FBgn0040505	Alk	Alk	ALK	Jelly belly	mtg, Ras1, rl
FBgn0053531	Ddr	Discoidin domain receptor	DDR1 and DDR2	Collagen?
FBgn0024245	dnt	Doughnut on 2	RYK	Wnt5?
FBgn0015380	drl	Derailed	RYK	Wnt5	Src64B
FBgn0033791	Drl-2	Derailed 2	RYK	Wnt5
FBgn0003731	Egfr	Epidermal growth factor receptor	EGFR	Spitz, Gurken, Vein, Keren	rho, Star, Ras1, Sos, csw, phl, Shc,dos, Gap1, Dsor, drk, ksr, cnk, rl, pnt, aop, ttk, sprouty, kekkon, argos
FBgn0025936	Eph/Dek	Eph receptor tyrosine kinase	EPHA and EPHB	Ephrin, Vap33, Exn	kuz, Exn, cac, Cdc42
FBgn0010389	htl/DFR1/Dtk1	Heartless	FGFR	Pyramus, Thisbe	Ras1, stumps, csw, rl, aop, pnt
FBgn0005592	btl/DFR2/Dtk2	Breathless	FGFR	Branchless	Ras1, stumps, csw, drk, Shc, Sos, ksr, cnk, sprouty, rl, grh, gro, pnt, aop
FBgn0013984	InR	Insulin-like receptor	INSR/IGF1R	Ilp1-7	chico, Sos, Drk, Shc, Ras, Pten, Pi3K92E, Pi3K21B, Pdk1, Tsc1, gigas, Rheb, Tor, Akt1, S6k, foxo
FBgn0038279	CG3837		INSR/IGF1R	Ilps?
FBgn0032752	CG10702		INSR/IGF1R	Ilps?
FBgn0032006	Pvr	PDGF- and VEGF-receptor related	VEGFR and PDGFR	PVF1,2,3	Ras1, rl, aop, Rac, mbc, ELMO, Crk, Cdc42
FBgn0011829	Ret	Ret oncogene	RET
FBgn0010407	Ror	One of two Ror kinases	Ror1 and Ror2	Orphan receptor
FBgn0020391	Nrk	Neurotropic receptor kinase	MuSK	Orphan receptor
FBgn0004839	otk/Dtrk	Offtrack	Trk	Wnt4	plexA, dsh
FBgn0003366	sev	Sevenless		Boss	Ras1, Sos, csw, phl, drk, dos, ksr, Gap1, Dsor, rl, aop, Pnt, Lz
FBgn0003733	tor	Torso		Trunk	Torso-like, fs(1)N, fs(1)ph, Ras1, Sos, csw, Shc, dos, Gap1, ksr, phl, Dsor, drk, rl, cic, gro
FBgn0022800	Cad96Ca/Stitcher	Cad96Ca			rl, grh
FBgn0014073	Tie	Tie-like receptor tyrosine kinase

Torso: AN RTK DETERMINANT OF ANTERIOR/POSTERIOR PATTERNING AND METAMORPHOSIS

The first RTK to be deployed during Drosophila embryogenesis is Torso. Torso is maternally contributed and localized uniformly to the membrane of the syncytial blastocyst. Localized activation of Torso involves the processing of its presumptive ligand, Trunk, at the egg poles, a process requiring at least three genes: torso-like, fs(1)Nasrat (fs(1)N), and fs(1)polehole (fs(1)ph) (Casanova and Struhl 1989; Sprenger et al. 1989; Stevens et al. 1990; Perrimon et al. 1995; Casali and Casanova 2001). Progeny derived from females lacking torso, trunk, or any of the aforementioned “terminal class genes” fail to develop stereotypical head and tail structures (Perrimon et al. 1986; Schupbach and Wieschaus 1986; Nüsslein-Volhard et al. 1987). Gain-of-function alleles of torso, on the other hand, drive the opposite phenotype: embryos with an extended posterior domain and minimal thoracic and abdominal regions (Klingler et al. 1988; Casanova and Struhl 1989; Schupbach and Wieschaus 1989; Strecker et al. 1989; Szabad et al. 1989). Screens to uncover suppressors of a torso gain-of-function allele identified Ras1 and son of sevenless (Sos). Further epistasis experiments positioned corkscrew (csw; SHP2), SHC-adaptor protein (Shc), GTPase-activating protein 1 (Gap1), kinase suppressor of ras (ksr), leonardo (leo; 14-3-3ζ), polehole (phl; RAF), Downstream of raf1 (Dsor; MEK), downstream of receptor kinases (drk; GRB2), and rolled (rl; ERK) within the hierarchy responsible for transducing the downstream signal from Torso (Ambrosio et al. 1989a,b; Casanova and Struhl 1989; Stevens et al. 1990; Perkins et al. 1992; Doyle and Bishop 1993; Lu et al. 1993, 1994; Tsuda et al. 1993; Brunner et al. 1994; Hou et al. 1995; Therrien et al. 1995; Li et al. 1997; Luschnig et al. 2000) (Fig. 1).

Figure 1.

Torso activation in embryogenesis. Processing of the Torso ligand Trunk occurs locally at the anterior and posterior embryonic poles and requires Torso-like, fs(1)N, and fs(1)ph. Engagement of Torso by processed Trunk triggers Torso autophosphorylation and subsequent recruitment of downstream adaptors and effectors. A phosphorylation cascade initiated by Tor activation and involving Raf/Phl, Mek/Dsor1, and ERK/Rl leads to the inhibition of transcriptional repression by Cic and Gro. This permits gap gene (tailless and huckebein) and subsequent pair-rule gene expression and enables patterning of the developing embryo.

Torso activation peaks between 1–2 hr of embryonic development (Sprenger and Nüsslein-Volhard 1992; Sprenger et al. 1993) results in the expression of tailless (tll) and huckebein (hkb), genes encoding for transcriptional repressors, at the embryonic poles (Moran and Jimenez 2006). These “terminal gap genes” demarcate zones of differentiation and embryos deficient for these gene products display phenotypes resembling those deficient for other members of the maternal terminal class (Pignoni et al. 1990; Weigel et al. 1990; Brönner and Jäckle 1991). The terminal class gene, rl/ERK, is upstream of tll, based on the fact that gain-of-function mutations in rl/ERK are unable to rescue tll null mutant embryos (Brunner et al. 1994). Phosphorylation of the transcriptional repressor Capicua (Cic) and corepressor Groucho (Gro) by activated ERK relieves transcriptional repression of tll and hkb at the embryonic posterior pole (Astigarraga et al. 2007; Cinnamon et al. 2008; Helman et al. 2011). At the anterior pole, Torso activation down-regulates the homeodomain transcription factor Bicoid by phosphorylation-independent mechanisms, in addition to inactivating Cic and Gro (Pignoni et al. 1992; Ronchi et al. 1993; Bellaïche et al. 1996; Janody et al. 2001). At both termini, Torso signaling inhibits Gro and permits Tll-dependent suppression of gap gene expression, whereas active Gro in central regions is unaffected and can repress tll expression, permitting central gap gene expression and in this way establishing expression “stripes” (Steingrímsson et al. 1991; Moran and Jimenez 2006). Gap genes encode for transcription factors that will activate expression of pair-rule genes. This sequential activation of gene expression enables patterning of the developing embryo (Nasiadka et al. 2002).

Torso also functions as a receptor for the neuropeptide prothoracicotropic hormone (PTTH) in the Drosophila brain during metamorphosis (Rewitz et al. 2009). Torso engagement by PTTH in the prothoracic gland (PG), an endocrine organ in insects, triggers Ras/Raf/ERK signaling to drive the production and/or release of the hormone ecdysone. Reduction of torso by RNAi specifically in the PG results in developmental delays similar to that resulting from ablation of PTTH-expressing neurons (McBrayer et al. 2007). PTTH shares significant structural homology with the Torso ligand Trunk and can substitute for Trunk in terminal signaling during embryogenesis (Rewitz et al. 2009).

Sevenless: AN RTK SPECIFYING CELL FATE IN THE Drosophila EYE AND TESTES

The Drosophila compound eye is comprised of 750–800 repetitive units termed ommatidia. Each ommatidium consists of eight photoreceptor neurons (R1–R8) and four lens-secreting cone cells, surrounded by a net of pigment cells that optically insulate each ommatidium from its neighbors (Wolff and Ready 1993). The spectral specificity of photoreceptor subtypes is provided by G-coupled Rhodopsin receptors. Photoreceptor differentiation occurs during the larval stage wherein a progressive “wave” of cell differentiation proceeds from the posterior to anterior region of the eye imaginal disc, the precursor of the adult eye. This wave (the morphogenetic furrow; MF) is visualized as a narrow indentation of epithelial cells contracting in the apical-basal dimension in a concerted fashion. Posterior to the MF, differentiated cells arrange into clusters and adopt the mature ommatidium pattern (Voas and Rebay 2003). A number of signaling pathways initiate MF progression and cell differentiation including epidermal growth factor receptor (EGFR), Notch, Wingless (Wg), Hedgehog (Hh), JAK-STAT, Decapentaplegic (Dpp), and Sevenless (Sev) (Charlton-Perkins et al. 2011).

R7 is the last photoreceptor to be recruited to the ommatidial cluster and specified. Differentiation of R7 relies on signals from neighboring cells within each ommatidial cluster; engagement of the RTK Sev on the surface of R7 by its membrane-associated ligand Bride-of-Sevenless (Boss), expressed exclusively in R8, triggers Ras/Raf/ERK signaling in R7 (Hart et al. 1990; Krämer et al. 1991; Simon et al. 1991). A lack of Sev activity in the R7 precursor (Tomlinson and Ready 1986, 1987; Tomlinson et al. 1987; Basler and Hafen 1988), or a lack of Boss in R8 (Reinke and Zipursky 1988), redirects R7 cell fate to that of a cone cell. Conversely, a cone cell precursor can be directed to become a R7 photoreceptor if the precursor expresses constitutively active Sev (Basler et al. 1991; Dickson et al. 1992; Sprenger and Nüsslein-Volhard 1992). Normally, the activation of Sev in photoreceptors other than R7, is restricted by the activity of Socs36E, expressed in all photoreceptors except R7 (Almudi et al. 2009, 2010), and reinforced in R7 by the adaptor protein Drk, specifically expressed in R7 (Olivier et al. 1993; Simon et al. 1993) (Fig. 2). Ras/Raf/Mek downstream from Sev (and EGFR—see below) results in the phosphorylation by Rl/ERK of two transcription factors critical for photoreceptor specification: Anterior open (Aop) and Pointed-P2 (Pnt-P2). Phosphorylation inhibits the repressor activity of Aop (O’Neill et al. 1994) by targeting it for nuclear export (Tootle et al. 2003) and degradation (Rebay and Rubin 1995). Pnt-P2, on the other hand, requires phosphorylation for its activity (Brunner et al. 1994; O’Neill et al. 1994). These factors play antagonistic roles in regulating the lozenge (lz) enhancer directly (Xu et al. 2000; Behan et al. 2002; Jackson Behan et al. 2005). High levels of RTK signaling in the prospective R7 cell relieve Tramtrack (Ttk)-dependent repression of lz (Daga et al. 1996; Xu et al. 2000; Siddall et al. 2009). Rl/ERK targets Ttk to the E3 ubiquitin ligase complex (comprised of Seven in absentia, Ebi, and Phyllopod) for degradation (Lai et al. 1997, 2002; Tang et al. 1997; Boulton et al. 2000; Li et al. 2002). Lz regulates prospero (pros) expression specifically in R7 by binding the pros enhancer directly. Lz works with Pnt-P2 at the pros enhancer, when Aop is displaced because of high ERK activity (Xu et al. 2000; Jackson Behan et al. 2005; Hayashi et al. 2008; Siddall et al. 2009). Pros functions to repress the expression of cone cell and R8-specific Rhodopsins thereby providing identity to R7 (Cook et al. 2003).

Figure 2.

Sevenless and EGFR signaling in photoreceptor specification. Engagement of Sev on the R7 photoreceptor cell by its ligand Boss on R8 activates the Ras/Raf/Mek/ERK signaling cascade in R7. A similar cassette of signaling proteins execute EGFR-dependent functions, following EGFR activation by its ligand Spitz. Sev activation in R8 is prevented by Socs36E and reinforced in R7 by the adaptor protein Drk, whereas Argos and Kekkon limit EGFR activation. Phosphorylated and active ERK/Rl targets the transcriptional repressors Aop and Ttk (via Phyl-Sina-Ebi) for degradation, whereas ERK-dependent phosphorylation stimulates Pnt transcriptional activity. This relieves transcriptional repression at lozenge and prospero enhancers. Lz functions together with Pnt to activate pros expression, thus providing R7 identity by repressing the expression of cone cell and R8-specific rhodopsins.

Genetic screening for modifiers of sev phenotypes identified many players downstream from Sev, including Ras1, Sos, Raf, Drk/Grb2, Dos, Csw, Gap1, and Rl/ERK (Rogge et al. 1991, 1992; Simon et al. 1991; Dickson et al. 1992; Gaul et al. 1992; Olivier et al. 1993; Biggs et al. 1994; Brunner et al. 1994; Raabe et al. 1996). These experiments were among the first to identify the genetic requirements for RTK signaling and in doing so delineate the hierarchy of canonical RTK signaling.

As is the case of Torso, Sev has also been shown to play a role in an additional cell type, in this case the male testes. Sev is required in a subset of somatic cells of the male embryonic gonad to spatially restrict stem cell niche differentiation. Sev activation in posterior somatic gonadal cells by Boss, presented by adjacent primordial germ cells, represses their differentiation into hub/niche cells and thereby restricts germline stem cell numbers (Kitadate et al. 2007).

EGFR: AN RTK WITH MULTITUDE ROLES

The Drosophila EGFR is involved in numerous developmental decisions throughout the Drosophila life cycle. For instance, the EGFR pathway has roles in dorsal/ventral patterning of the embryonic ectoderm and the establishment of neuroectoderm, wing development, antennal formation, photoreceptor differentiation, lamina neuron differentiation, and the specification of muscle precursors and invagination of tracheal branches to name a few (Perrimon and Perkins 1997; Shilo 2003). EGFR predominantly mediates short-range signaling that is restricted either to the cells producing EGF or to cells positioned 1–2 cells away. EGFR activates the canonical Ras/Raf/MEK/ERK pathway and visualization of pathway activity has documented the highly dynamic activity of EGFR signaling (Gabay et al. 1997) (Fig. 2). Multiple EGFR ligands and feedback loops are responsible for the complex temporal and spatial regulation of EGFR signaling (Perrimon and McMahon 1999; Shilo 2005). Regarding activation, there are four EGFR ligands in Drosophila: Spitz, Keren, Gurken, and Vein, with Vein being the only secreted protein that does not require processing for its activity. Detailed studies of ligand processing, in particular of Spitz, have shown a requirement for two proteins, Star and Rhomboid (Rho). Star is a type II transmembrane protein that associates with Spitz, facilitating its translocation from the ER to a cellular compartment where it is cleaved by the seven-pass transmembrane protein Rhomboid (Rho). Importantly, although Star expression is ubiquitous, Rho is extremely dynamic and thus responsible for controlling EGFR activation in a wide range of tissues. Finally, the transcriptional induction of negative regulators of the pathway restricts the spatial and temporal activation of EGFR signaling. These include the cytoplasmic proteins Sprouty and Cbl, the extracellular secreted molecule Argos, and the extracellular transmembrane protein Kekkon. A comprehensive description of the entirety of EGFR function during Drosophila development would exhaust the page limitations of this chapter and so we refer readers to a comprehensive review (Shilo 2005).

FGFR SUPERFAMILY: HEARTLESS AND BREATHLESS

Heartless: An RTK Influencing Mesodermal Cell Migration and Cell Specification

The gene product encoded by heartless (htl) shares ∼60% identity in its kinase domain with vertebrate FGFRs and 80% identity with the other FGFR homolog in Drosophila, Breathless (Btl). htl is expressed ∼2.5 hr postfertilization in presumptive mesoderm at the onset of gastrulation. htl expression persists throughout embryogenesis in somatic muscle precursors including cardiac and pericardial cells, pharyngeal muscle cells, visceral muscle precursors, and additionally in glia of the central nervous system (CNS) (Shishido et al. 1993; Hidalgo and Booth 2000; Egger et al. 2002; Freeman et al. 2003). During larval stages, htl is expressed in muscle cell precursors of wing and leg imaginal discs and in neural precursors and glia of the brain and eye imaginal discs (Emori and Saigo 1993; Sato and Kornberg 2002; Butler 2003; Butler et al. 2003; Franzdóttir et al. 2009). During pupal and adult stages, htl is expressed in abdominal and thoracic myoblasts (Dutta et al. 2005).

After ventral furrow invagination, the mesoderm primordium undergoes an epithelial-to-mesenchymal transition and spreads dorsally over ectodermal cells to form a monolayer (Schumacher et al. 2004; Wilson 2005; Wilson et al. 2005; Clark et al. 2011). This repositioning is required for the reception by mesodermal cells of patterning cues (Dpp and Wg) from the adjacent ectoderm that specifies the mesoderm lineage into visceral mesoderm, heart tissue, somatic muscle, and the fat body (FB). htl null mutant embryos show defects in the bilateral spreading of mesoderm during gastrulation (Murray and Saint 2007; McMahon et al. 2008). htl mutants fail to develop visceral mesoderm and heart tissue, whereas somatic muscles are disorganized and reduced because of the absence of differentiated mesodermal subtypes. htl mutants show additional defects: failure of CNS glia to migrate and ensheath longitudinal ventral nerve cord (VNC) connectives, and defective salivary gland migration (Beiman et al. 1996; Gisselbrecht et al. 1996; Shishido et al. 1997; Michelson et al. 1998b; Schulz and Gajewski 1999; Mandal et al. 2004).

Expression of activated Ras1 partially rescues mesodermal defects associated with htl perturbation (Beiman et al. 1996; Gisselbrecht et al. 1996; Michelson et al. 1998b; Schulz and Gajewski 1999). Ras1 functions downstream or parallel to the adaptor protein Stumps (Carmena et al. 1998; Michelson et al. 1998a; Vincent et al. 1998; Imam et al. 1999) to transduce signals downstream from Htl (Johnson Hamlet and Perkins 2001; Petit et al. 2004; Csiszar et al. 2010). Ectopic expression of activated Aop, a transcriptional repressor downstream from ERK, generates phenotypes similar to that due to htl disruption whereas expression of activated Pnt, a transcriptional activator downstream from ERK, increases the number of somatic muscle progenitors (Halfon et al. 2000). These observations are consistent with Htl-dependent ERK activation (Gabay et al. 1997; Wilson et al. 2004).

Thisbe (ths) and pyramus (pyr), ligands for Htl, are expressed in the neurogenic ectoderm coincident with the migration of mesoderm during gastrulation. They are later differentially expressed in other epithelial tissues that flank mesodermal derivatives: the stomadeum, the hindgut, the CNS, and at muscle attachment sites. ths and pyr mutants are defective in mesodermal cell intercalation and monolayer formation after dorsal spreading. ERK activation at the leading edge of the migrating mesoderm is absent in ths pyr double mutants, and expanded as a result of ectopic ths expression similar to that attributable to constitutively active Htl expression. Constitutively active Htl partially restores mesodermal differentiation to ths pyr mutants (Gryzik and Müller 2004; Stathopoulos et al. 2004; Klingseisen et al. 2009; McMahon et al. 2010; Clark et al. 2011).

The anterior migration of caudal visceral mesoderm, giving rise to the longitudinal muscles that ensheath the gut, is guided by Htl activation. ths and pyr, expressed in adjacent trunk visceral mesoderm, together promote cell survival and restrict lateral movement of caudal visceral mesoderm cells during their migration along trunk visceral mesoderm (Kadam et al. 2012). In the cardiogenic mesoderm, Pyr plays the major role in activating ERK to maintain cardiogenic lineages (Klingseisen et al. 2009; Grigorian et al. 2011). In the eye imaginal disc, ths and pyr have different expression patterns that translate to unique contributions to Htl signaling: pyr for early glia–glia interactions that promote glial cell proliferation and migration, and ths for glial–neuron interactions that inhibit migration and trigger cell differentiation (Franzdóttir et al. 2009).

Htl is additionally required in the gonadal mesoderm for primordial germ cell (PMC) migration. In htl mutant embryos, PMCs transverse the posterior midgut but stall at the endoderm/lateral mesoderm border. Those few PMCs that infiltrate the lateral mesoderm fail to navigate toward and associate with somatic gonadal precursors (SGPs)—specialized mesodermal cells that give rise to the somatic portion of the gonad. SGPs of htl mutant embryos are reduced in number and deranged in shape (Moore et al. 1998).

htl is required in the Drosophila ocellar sensory system (OSS), to direct OSS axon development during pupariation. OSS axons migrate toward their targets in the brain, until metamorphosis when they become detached and reorient. Properties such as the ability to attach, detach, or cross to the brain are lost when dominant-negative Htl is expressed in neurons. Genetic evidence implies Htl functions downstream from Neuroglian—a homophilic cell adhesion molecule required for axon guidance—in this context (García-Alonso et al. 2000).

Htl is necessary for larval cardiac tube remodeling, which occurs without cell migration. Htl is additionally required for the formation of ventral imaginal muscle founders and the differentiation of leg imaginal disc associated myoblasts and abdominal/thoracic adult myoblasts. Modulation of Htl activity alters the number of myoblast founder cells and adult muscle fibers (Dutta et al. 2005; Maqbool et al. 2006; Zeitouni et al. 2007).

Breathless: An RTK Influencing Cell Migration and Patterning, Predominantly in the Tracheal System and CNS

Although they share significant identity, the two FGF receptors in Drosophila, Breathless (Btl) and Htl differ in their ligand binding domain structure (Shishido et al. 1993). As such, the Htl ligands Pyr and Ths are unable to activate Btl to influence tracheal branching whereas the Btl ligand, Branchless (Bnl), is unable to influence Htl-dependent mesoderm spreading and differentiation (Kadam et al. 2009).

btl is expressed during embryogenesis in the invaginating tracheal primordia and developing tracheal system, the salivary glands, CNS glia and neurons, cells of the gut and male genitalia primordium (Glazer and Shilo 1991; Klambt et al. 1992; Shishido et al. 1993; Ahmad and Baker 2002). bnl is expressed in cells surrounding Btl-expressing cells and prefigures their migratory direction (Sutherland et al. 1996).

btl and bnl mutant embryos show defects in tracheal cell migration; however, the specification and proliferation of tracheal precursors is normal. The absence of a tracheal system skeleton in btl mutants ensues from an inability of cells to coordinately migrate out from the tracheal placodes in stereotyped directions and then intercalate and elongate to form tubes. Further, the formation and fusion of secondary and terminal branches, each derived from a single cell, is compromised in btl mutants because of a failure of tracheal cell fate acquisition. Proper tracheal branching relies on the spatial regulation of Btl activity; localized misexpression of bnl can redirect tracheal cell migration and induce branching, through the activation of Btl and a downstream Pnt-dependent gene expression program (Klambt et al. 1992; Reichman-Fried and Shilo 1995; Lee et al. 1996; Samakovlis et al. 1996; Sutherland et al. 1996). In this same manner, oxygen deprivation directs fine terminal branching during larval stages—triggering bnl expression and therefore, Btl activation—for oxygen delivery (Jarecki et al. 1999; Ghabrial et al. 2011). Branching relies on an extensive number of factors downstream from Btl (Ghabrial et al. 2011). For instance, Btl autophosphorylation, following Bnl binding (Lee et al. 1996; Sutherland et al. 1996), functions to recruit the adaptor protein Stumps. Phosphorylation of Stumps by Btl induces binding of the phosphatase Csw, another component of the signaling cascade that activates ERK (Michelson et al. 1998a; Vincent et al. 1998; Imam et al. 1999; Wilson et al. 2004) (Fig. 3). ERK function is discharged by the transcriptional activator Grh (Hemphala et al. 2003) and the transcriptional corepressor Gro (Cinnamon et al. 2008). ERK-dependent phosphorylation of these factors as well as of the transcription factor Aop, a repressor of btl, modulates their activity (Ohshiro et al. 2002). The FGFR inhibitor Sprouty limits the range of Bnl signaling and prevents tracheal branch stalk cells from budding ectopically (Hacohen et al. 1998).

Figure 3.

Breathless signaling in the tracheal system. Btl autophosphorylation, on Bnl binding, recruits the adaptor protein Stumps and additional downstream effectors. Ras activation initiates a phosphorylation cascade that culminates in the stimulation of ERK kinase activity. ERK-dependent phosphorylation of the transcriptional activators Grh and Pnt, and the transcriptional repressors Gro and Aop, modulates their activity at promoters. The consequential up-regulation of gene expression induces tracheal cell migration and tracheal branching and fine-tuning of Btl signaling.

Tracheal branch fusion relies on Btl-dependent Delta expression in fusion cells of migrating branches. Delta displayed by the fusion cell activates Notch on adjacent cells to limit fusion cell identity to one cell per branch; Notch down-regulates bnl expression to restrict fusion cell identity and delimit ERK activation (Ikeya and Hayashi 1999). Notch-mediated lateral inhibition also restricts the number of leading cells in a branch (Ghabrial and Krasnow 2006). Fusion further depends on a single mesodermal cell—the bridge cell—that guides the extension and fusion of tracheal metameres to generate a continuous dorsal trunk (Wolf and Schuh 2000). Bnl induces filopodial tracheal cell extensions that contact the bridge cell and are essential for dorsal trunk branch fusion (Ribeiro et al. 2002; Wolf et al. 2002).

Btl/Bnl-dependent filopodial extensions are also presented by imaginal tracheoblasts, which proliferate to remodel the adult respiratory system during metamorphosis. The migration of tracheoblasts at the tip of the air sac primordium is dependent on Btl (Sato and Kornberg 2002; Cabernard and Affolter 2005). Btl additionally influences tracheoblast identity in the spiracular branch and the dorsal branch stalk, by inducing tracheoblast differentiation as well as by promoting differentiated cells to reenter the cell cycle, respectively (Guha et al. 2008; Sato et al. 2008; Weaver and Krasnow 2008; Pitsouli and Perrimon 2010).

btl-expressing cells migrate toward distinct populations of bnl-expressing cells additionally in the male genital imaginal disc; bnl expression by ectodermally-derived genital precursor cells induces btl-expressing cell migration into the male disc. The btl-expressing mesodermal cells are subsequently converted into epithelia during pupal stages to generate the vascular paragonia and vas deferens. In female genital discs, bnl expression is targeted by the female-specific repressor form of the transcription factor Doublesex (Ahmad and Baker 2002).

In the embryonic CNS of btl mutants, posterior midline glial cells migrate inappropriately, resulting in irregular commissural patterning (Klambt et al. 1992). Notably, perturbation of pnt and stumps also generates glial cell migration phenotypes (Klambt 1993; Vincent et al. 1998; Imam et al. 1999). During larval stages of eye patterning Bnl is required for cell adhesion and Hh-dependent apical constriction that enables ommatidial cluster formation. Btl is further required cell autonomously for retinal architectural integrity and noncell-autonomously in directing retinal glia migration (Mukherjee et al. 2012). In the adult brain, Btl is required to mediate axonal retraction rather than guidance specifically in the dorsal cluster neurons (DCN) of the visual system. In this context, Btl is activated in extending DCN axons as they encounter Bnl. Btl signaling activates Rac that in turn inhibits JNK signaling, inducing axonal retraction (Srahna et al. 2006).

INSULIN RECEPTOR RTK SUPERFAMILY: THE INSULIN RECEPTOR AND ANAPLASTIC LYMPHOMA KINASE

The Insulin and Insulin-Like Growth Factor Receptor in Drosophila: An RTK Essential for Growth

The insulin-like growth factor receptor in Drosophila, InR, is ubiquitously expressed throughout embryogenesis, with higher levels accumulating in the brain, midgut primordia, and VNC. A maternally inherited role for InR is reflected by abundant InR mRNA in nurse cells and mature oocytes (Petruzzelli et al. 1986; Garofalo and Rosen 1988). Accordingly, embryonic lethality is associated with InR complete loss-of-function mutations whereas some heteroallelic combinations yield animals that are viable but sterile (Fernandez et al. 1995; Chen et al. 1996; Tatar et al. 2001). Viable embryos lack neuroblasts and are unable to complete germband retraction and dorsal closure. They display abnormal head structures and cuticle. Moreover, mutant embryos display defects in VNC condensation and commissure formation. InR expression persists in the nervous system during larval stages, and is detected in all imaginal discs and postsynaptically at neuromuscular junctions (NMJs) (Garofalo and Rosen 1988; Gorczyca et al. 1993; Fernandez et al. 1995). InR is enriched in photoreceptor axons of late larvae (Song et al. 2003). In adults, InR mRNA is predominantly localized to the brain cortex, cells of the thoracic and abdominal ganglia, and the gut (Veenstra et al. 2008).

Small animals result from reduced InR activity (Fernandez et al. 1995; Chen et al. 1996; Brogiolo et al. 2001; Tatar et al. 2001). Organismal size is additionally affected by alteration of conserved InR signaling pathway components. Positive regulators downstream from InR include: the insulin receptor substrate (IRS) ortholog Chico (Böhni et al. 1999), the PI3K subunits PI3K92E and PI3K21B (Weinkove et al. 1999), the PI-dependent protein kinase Pdk1 (Rintelen et al. 2001), and the kinases Tor (Oldham et al. 2000), Akt1 (Verdu et al. 1999), and S6K (Montagne et al. 1999) (Fig. 4). Negative regulators downstream from InR include: the phosphatase Pten (Gao et al. 2000; Oldham et al. 2002), the tuberous sclerosis genes Tsc1 and Gigas/Tsc2 (Gao and Pan 2001; Potter et al. 2001), and the transcription factor Foxo (Jünger et al. 2003). Cell-autonomous effects of InR on cell size and number rely on kinase activity (Brogiolo et al. 2001) and are antagonized by Pten overexpression (Huang et al. 1999), co-expression of Tsc1 and gigas (Potter et al. 2001), or reduced Pdk1 activity (Rintelen et al. 2001). Although PI3K92E or Akt1 overexpression increases cell size it fails to influence cell number or division (Verdu et al. 1999; Weinkove et al. 1999), substantiating the supposition that InR serves two independent functions: to promote proliferation via Ras/ERK and to promote protein synthesis through PI3K (Oldham et al. 2002). The list of genes contributing to body size determination and functioning downstream from InR continues to grow and to date includes numerous conserved and Drosophila-specific factors.

Figure 4.

Ubiquitous InR signaling. Interaction of InR with its Ilp ligand induces InR autophosphorylation. Chico interacts directly with the phosphorylated receptor, recruiting PI3K along with a multitude of other factors including adaptor proteins and downstream effectors. InR promotes proliferation via canonical Ras/Raf/Mek/ERK signaling and growth via the activation of Akt1. Akt1 stimulates protein synthesis by activating downstream kinases Tor and S6K, and by inhibiting Foxo nuclear accumulation and transcriptional activity.

Seven insulin-like peptides (Ilp1-7), InR ligands, are differentially expressed among Drosophila developmental stages and tissues: Ilp2, Ilp4, and Ilp7 are expressed in the mesoderm and midgut; Ilp1, Ilp2, Ilp3, and Ilp5 are expressed in neurosecretory cells of the brain that project to the ring gland, heart and brain lobes, and foregut; Ilp2 is expressed in imaginal discs and salivary glands; Ilp7 is expressed in neurons of the ventral ganglion, a subset of which innervates the adult hindgut and another that makes synaptic contact with Ilp1,2,3,5-expressing neurosecretory cells; Ilp5 is expressed in ovarian follicle cells; Ilp3 is expressed in adult midgut muscle; and Ilp6 is expressed in the FB, gut, and in CNS surface glia (Gorczyca et al. 1993; Brogiolo et al. 2001; Ikeya et al. 2002; Rulifson et al. 2002; Broughton et al. 2005; Miguel-Aliaga et al. 2008; Veenstra et al. 2008; Okamoto et al. 2009; Slaidina et al. 2009; Chell and Brand 2010; Cognigni et al. 2011). Ilp overexpression increases organismal size by stimulating cell growth and division (Brogiolo et al. 2001; Ikeya et al. 2002; Slaidina et al. 2009). FB-specific activation of InR promotes triglyceride storage, by increasing fat cell number and lipid content (DiAngelo and Birnbaum 2009). Reciprocally, ablation of Drosophila Ilp-expressing neurons causes developmental delays, a reduction in egg production rates, and a reduction in organismal size owing to decreased cell number and size (Ikeya et al. 2002; Rulifson et al. 2002; Broughton et al. 2005). In addition to developmental disparities, flies devoid of Ilp-producing neurons have elevated levels of glucose and trehalose, like InR and chico mutants.

Although the InR-signaling pathway autonomously influences growth and proliferation, it nonautonomously influences aging. Long-lived animals result from reduced activity of InR, Chico, Tor, and S6K (Clancy et al. 2001; Tatar et al. 2001; Kapahi et al. 2004). Although nutritional starvation during larval stages influences organismal size and fecundity, it is insufficient to influence aging (Tu and Tatar 2003). Dietary restriction in adults, however, promotes longevity and correlates with reduced Ilp5 mRNA levels (Min et al. 2008). Ablation of Ilp-producing neurosecretory cells can also extend lifespan (Broughton et al. 2005).

Recently, both the reactivation of proliferation in quiescent embryonic neuroblasts and the elimination of larval neuroblasts postpupariation were shown to correlate with InR signaling. The reception by quiescent embryonic neuroblasts of “nutritional adequacy” signals from the FB elicits Ilp expression in glial cells, activating InR and proliferation in quiescent neuroblasts in a paracrine manner (Britton and Edgar 1998). This reactivation of InR relies on Tor and the amino acid transporter Slimfast (Slif) in the FB, and functional InR/PI3K/Akt/Tor in neuroblasts. Consistent with InR promoting proliferation and survival, InR activation in combination with proapoptotic gene ablation delays larval neuroblast elimination. Reduced InR signaling, on the other hand, promotes larval neuroblast elimination postpupariation. In light of observations wherein impaired autophagy enhances caspase-dependent neuroblast survival, Foxo is proposed to activate autophagy in aging neuroblasts (Chell and Brand 2010; Siegrist et al. 2010; Sousa-Nunes et al. 2011). This is consistent with the acknowledged antagonism between autophagy and insulin signaling pathways (Chang et al. 2009) and the proposed mechanism of lifespan extension by Foxo in muscle (Demontis and Perrimon 2010). It is plausible that two other genes in the Drosophila genome, CG3837 and CG10702, predicted to encode for insulin-like receptors (Table 1), play a role in this context given that elevated levels of the corresponding proteins have been detected in salivary glands undergoing autophagic programmed cell death (Martin et al. 2007).

In addition to growth and longevity, InR also impinges on reproduction. InR mutant eggs chambers, unable to interpret both follicular cell-derived and systemic insulin signals, are developmentally delayed owing to impaired vitellogenesis and cyst growth. InR, deficient germline stem cells (GSCs) show reduced division rates owing to compromised G₂ phase of the GSC division cycle, whereas hindered GSC division in InR mutant females reflects additional effects on Notch activity and E-cadherin-mediated adhesion within the niche (Drummond-Barbosa and Spradling 2001; LaFever and Drummond-Barbosa 2005; Hsu et al. 2008; Yu et al. 2009; Hsu and Drummond-Barbosa 2011). An analogous scenario exists in the male germline: insulin signaling in germ and support cells promotes spermatocyte growth and maintains cyst numbers by promoting G₂/M phase progression of GSCs (Ueishi et al. 2009; McLeod et al. 2010).

Ilps are expressed in synaptic boutons at the presynaptic terminals of larval body wall muscles (Gorczyca et al. 1993). Overexpression of PI3K92E, Akt or InR, specifically in larval motor neurons, induces supernumerary synapses projecting to each body wall muscle whereas reduced PI3K or Akt activity reduces synapse number. Similar effects are seen in projection neurons of the brain. Alterations of PI3K activity in motor and projection neurons elicits modified locomotive behaviors (Martin-Pena et al. 2006).

InR mutants display defects in photoreceptor axon path finding reminiscent of animals deficient for Dystroglycan and Dystrophin—two genes linked to muscular dystrophies. InR interacts genetically with Dystroglycan and is speculated to function independently of Chico and the PI3K/Akt pathway in this context, and instead associates with the Dg-Dock-Pak pathway to guide neuronal migration (Song et al. 2003; Shcherbata et al. 2007).

Alk: An RTK Involved in the Development of the Visceral Mesoderm, as Well as Motor and Visual Circuitry

Alk is required for the generation of visceral mesoderm—cells that will comprise the inner circular muscles and outer longitudinal muscles that ensheath the intestinal tract. Alk deficient embryos are unable to specify visceral mesoderm founder cells and completely lack visceral musculature. Alk mutants lack intestinal structures, do not eat and die at the first instar larval stage (Englund et al. 2003; Lee et al. 2003; Lorén et al. 2003).

The derivation of mesoderm in Drosophila relies on the coordinated activity of Dpp, Hedgehog (Hh) and Wg. Signaling in these pathways is mediated by Tinman, a conserved homeodomain transcription factor required for somatic, cardiac and visceral mesoderm development (Azpiazu and Frasch 1993; Bodmer 1993; Furlong 2004). A screen for Tinman targets identified jelly belly (jeb). jeb is expressed in early ventral and medial somatic mesoderm cells, adjacent to Alk-expressing visceral mesoderm cells. Only visceral mesoderm is affected in jeb mutants; ergo, the logical gene name—jelly belly—referring to a jiggly expansive abdomen and lack of midgut muscles (J Weiss, pers. comm.). The jeb mutant phenotype is akin to that of Alk and is reflective of a failure of visceral mesodermal cells to differentiate and migrate, because of a lack of either Jeb secretion by ventral somatic mesoderm precursor cells or Jeb engagement by Alk at visceral muscle precursors (Weiss et al. 2001; Stute et al. 2004). Engagement of Alk by Jeb activates ERK in visceral muscle progenitors to establish two separate pools of cells: founder myoblasts and fusion-competent myoblasts. In the absence of jeb or Alk, all visceral mesoderm progenitors become fusion-competent myoblasts and these cells fuse with the somatic muscle founders because founder myoblasts are nonexistent, resulting in a complete lack of visceral musculature (Stute et al. 2004). Ectopic expression of activated Alk in the visceral mesoderm restores gut morphogenesis to jeb mutants by reinstating expression of genes downstream from ERK: dumbfounded, org-1, sticks and stones, Hand, and Dpp (Englund et al. 2003; Lee et al. 2003; Stute et al. 2004; Varshney and Palmer 2006; Shirinian et al. 2007).

A proposed function for Alk in neuronal development is based on conserved nervous system expression in C. elegans, mouse, and chick (Iwahara et al. 1997; Liao et al. 2004; Hurley et al. 2006; Vernersson et al. 2006; Reiner et al. 2008). Like Alk, jeb is expressed in a subset of neurons distributed throughout the VNC in late Drosophila embryogenesis. The accumulation of Jeb in CNS axons is dependent on Alk (Englund et al. 2003; Lee et al. 2003; Stute et al. 2004; Rohrbough and Broadie 2010). Further, postsynaptic Jeb internalization and Alk accumulation at developing NMJs during embryonic and larval stages is regulated by Mind the gap (Mtg), a neuronally secreted glycoprotein required for synaptic cleft extracellular matrix assembly. Rescue of defective jeb mutant larvae locomotion by neuronal-specific jeb expression unveiled a neuronal requirement for Jeb (Rohrbough and Broadie 2010).

During pupal stages, Alk functions in the visual system in photoreceptor axon target selection. jeb is expressed in photoreceptor axons, whereas Alk is expressed in processes of target neurons innervating the optic lobe. Photoreceptor axons secrete Jeb to activate Alk in target neurons within the lamina and medulla. R7 and R8 photoreceptor axons are significantly altered when Alk is removed from target neurons (Bazigou et al. 2007). jeb mosaic animals also show photoreceptor projection targeting defects. These photoreceptor navigation errors are reflective of a lack of regular lamina cartridge patterning and defective duf expression (Englund et al. 2003; Lee et al. 2003; Stute et al. 2004).

Olfactory learning deficits are associated with Alk activation and surprisingly elevated performance correlates with Alk inactivation. In addition to impaired learning, animals bearing constitutive Alk activation are small and conversely animals deficient for Alk activity are large. These nonautonomous effects on organismal size rely on the activation of Alk in peptidergic and cholinergic neurons and are insulin independent. Alk displays genetic interactions with the Ras GTPase activating protein Nf1­. Consistent with the genetics, Alk activates ERK whereas Nf1 functions as a negative regulator of ERK (Walker et al. 2006). These antagonistic roles for Alk and Nf1 are proposed to regulate the release of GABA neurotransmitter and account for learning and long-term potentiation phenotypes (Ho et al. 2007; Liu et al. 2007).

Alk plays a vital role in protecting the developing CNS from nutrient deprivation. Under these conditions, flies eclose with relatively small bodies. Their heads, however, are of normal size. This is because nutrient deprivation triggers Jeb secretion from glia, activating Alk in the neural progenitors of the brain. Alk stimulates the activation of the InR effectors S6K, Thor, and PI3K, specifically in the brain, salvaging it from nutrient restriction (Cheng et al. 2011). This mechanism is Ras independent and therefore distinct from that described for Alk elsewhere. Of all Drosophila RTKs, Alk shares the most sequence similarity with InR. Mammalian ALK binds IRS-1 and SHC via its NPXpY motif. This motif exists in Drosophila Alk and offers a mechanism by which Alk activates PI3K when nutrients are limiting (Fujimoto et al. 1996).

THE TRK SUPERFAMILY OF RTKs: Ror, Nrk, AND DDR

Ror: A Drosophila Orphan RTK with Roles Exclusively in the Nervous System

Although Drosophila Ror is most similar to human Ror1 and Ror2 (∼35% identity), it lacks many of the domains found in the human Ror RTKs. Ror expression begins ∼6.5 hr after fertilization and peaks between 8–12 hr. Ror is expressed exclusively in neurons of the CNS and PNS (Wilson et al. 1993). This time in Drosophila development coincides with neural differentiation and axonogenesis and as such Ror is suggested to be involved in these processes. Consistent with this, the C. elegans Ror ortholog CAM-1 regulates neuronal polarity and the asymmetric division of neurons (Forrester et al. 1999) and vertebrate Rors play roles in neurite outgrowth and synapse formation (Paganoni and Ferreira 2005; Paganoni et al. 2010). A mutant phenotype for Ror in Drosophila, however, has not been described.

Ror shows a pattern of cysteine residue spacing suggestive of ligand binding like that for other transmembrane receptors such as Frizzled (Fz) (Saldanha et al. 1998). Like Fz, this region binds Wnt ligands in C. elegans, Xenopus, and mammalian cell lines (Hikasa et al. 2002; Oishi et al. 2003; Mikels and Nusse 2006; Green et al. 2007). The interaction of Ror proteins with various Wnts has been implicated in diverse contexts including: the migration and asymmetric division of neurons and vulval precursor cells in C. elegans, convergent extension in Xenopus, mouse embryonic fibroblast migration, and synapse formation in the mouse brain (Green et al. 2008; Grumolato et al. 2010; Paganoni et al. 2010). To date, Ror binding to Wnt has not been reported in Drosophila.

Neurotropic Receptor Kinase: Another Drosophila Orphan RTK in the Nervous System

Neurotropic receptor kinase (Nrk) is considered a MuSK ortholog, based on extensive homology in its kinase domain (Sossin 2006). Like Ror RTKs, MuSK binds Wnt ligands via the Fz domain (Jing et al. 2009). Although differences exist between Nrk and MuSK extracellular domains, Nrk has an Fz domain and is predicted to bind Wnts. In vertebrates, MuSK induces acetylcholine receptor clustering at NMJs and the stability of clusters relies on the heparan–sulphate proteoglycan Agrin and LRP coreceptor (DeChiara et al. 1996; Bezakova et al. 2001; Kim et al. 2008; Zhang et al. 2008). Nrk, however, lacks the extracellular domains responsible for Agrin binding as well as the intracellular NPXpY motif in MuSK essential for NMJ formation (Herbst and Burden 2000). MuSK has an Agrin-independent role in axon guidance in zebrafish (Zhang et al. 2004) and a noncanonical Wnt-mediated function in neural crest cell migration in mouse (Banerjee et al. 2011). These studies hint at similar roles for Nrk.

Nrk expression begins ∼9.5 hr after fertilization, shortly after the determination of neural precursor cells. Expression of Nrk is initially detected in the neuroectoderm and becomes restricted to neural progenitor cells situated between epidermal and mesodermal cell layers. Expression persists in the neural cell lineage throughout embryogenesis and peaks again at the pupal stage during restructuring of the nervous system (Oishi et al. 1997). In support of a role for Nrk in the nervous systems, Nrk expression was down-regulated in embryos for which neuroectoderm was derived primarily from glial cells, rather than both neurons and glia (Egger et al. 2002).

Discoidin Domain Receptor: A Poorly Characterized RTK in Drosophila

Discoidin domain receptors (DDR) are atypical RTKs in that they are activated by collagen rather than secreted factors. Moreover, maximal activation of DDRs occurs several hours after collagen binding, unlike other RTKs that are activated within minutes of receptor engagement (Shrivastava et al. 1997; Vogel et al. 1997). Although Drosophila Ddr has not yet been shown to bind collagen, conservation of the discoidin domain suggests that it should possess this ability (Sossin 2006). DDRs have been implicated in cell migration, extracellular matrix remodeling, proliferation and differentiation in a number of tissues including vertebrate lung, skin, GI tract, kidney, heart, liver, mammary gland, endometrium, and brain (Vogel et al. 2006). Like its mammalian counterparts, the sole Drosophila Ddr bears an extensive number of tyrosine residues in its cytoplasmic region. In mammals, these residues recruit a number of downstream factors following phosphorylation including Shc, Nck2, Shp-2, PI3K, RasGAP, and Stats (Lemeer et al. 2011). Wnt5a was shown to be required for collagen-induced activation of DDR in a breast cancer cell line (Jonsson and Andersson 2001). It is tempting to speculate that Wnt5A might function as a ligand for DDR, in a manner analogous to that of Ror and Ryk RTKs.

OFFTRACK: A “DEAD” RTK WITH ROLES IN MOTOR AND CNS AXON TARGETING AND EMBRYONIC PATTERNING

Drosophila off-track (also known as Dtrk) shares 65% similarity with human Trk. off-track (otk) is expressed 3–4 hr postfertilization in the anterior midgut primordia, the cephalic furrow, and along the germ band. Expression peaks mid-embryogenesis in neuroectodermal cells and internalized CNS neuroblasts. Otk expression persists in prospective gut and head regions, and eventually accumulates throughout the CNS in segmentally repeated commissures, in axon bundles exiting the CNS, in motor neuron projections innervating muscle fibers, and in neurons of a subset of sensory organs. Notably, otk null animals are embryonic lethal (Pulido et al. 1992; Winberg et al. 2001).

Otk bears structural similarity with cell adhesion molecules of the immunoglobulin (Ig) superfamily expressed in the Drosophila nervous system. Like other Ig superfamily members, Otk is glycosylated and permits cell aggregation in vitro (Pulido et al. 1992). This property likely influences neuroblast migration and axon targeting. Accordingly, Otk is implicated in Sema-1a-mediated embryonic motor and CNS axon guidance based on: (1) its physical association with the repulsive axon guidance receptor Plexin A; (2) impaired defasciculation and disrupted axon morphology and targeting in otk mutant embryos, similar to that resulting from Sema-1a or PlexA inactivation; and (3) genetic interactions between otk and PlexA or Sema-1a (Pulido et al. 1992; Winberg et al. 2001). Conserved catalytic residues in the kinase domain are altered in Otk and as such Otk belongs to the CCK-4 subfamily of “dead” RTKs (Kroiher et al. 2001). Otk itself is tyrosine phosphorylated, which may serve to recruit signaling molecules to a Sema-1a-PlexA-Otk complex (Pulido et al. 1992; Winberg et al. 2001).

A screen for R1–R6 photoreceptor growth cone targeting identified otk; an increased number of R1–R6 photoreceptor axons project through the lamina and inappropriately into the medulla of the developing larval optic lobe in otk mosaic heads. The role of Otk in guiding R1–R6 axons appears to be unrelated to that of Otk in Sema-1a signaling (Cafferty et al. 2004). The Otk ligand responsible for R1–R6 targeting is unknown.

Recently, Otk binding to Wnt4 was shown to inhibit canonical β-catenin/TCF signaling in an Fz-dependent but LRP-independent manner. Both otk and Wnt4 mutants show embryonic patterning defects indicative of excessive canonical β-catenin/TCF signaling. In the ventral embryonic epidermis and the adult wing, otk and Wnt4 overexpression reduce canonical β-catenin/TCF signaling. Ectopic Wnt4 expression phenotypes rely on functional Otk and synergizes with otk overexpression. Otk is postulated to direct noncanonical Wnt signaling through interaction with Disheveled, and inhibit canonical Wnt signaling by either occluding LRP or by sequestering canonical Wnts (Peradziryi et al. 2011).

THE RYK SUPERFAMILY OF RTKs: DERAILED AND DOUGHNUT

Derailed: Another “Dead” RTK with Roles in Neuronal Pathway Selection and Muscle Attachment

derailed (drl) expression begins ∼6 hr postfertilization, in the embryonic epidermis and in salivary placodes. By 10 hr, Drl is detected in somatic muscle 21–23 precursors and their associated epidermal cells. By 12 hr, drl expression is detected in a subset of heterogeneous neurons that project in the anterior commissure (AC) of the VNC (Callahan et al. 1995, 1996; Bonkowsky and Thomas 1999; Harris and Beckendorf 2007). drl is expressed later in the larval CNS, as well as the adult brain. Homozygous drl mutant animals are viable but uncoordinated because of drl axon mistargeting and defasciculation, and the inability of muscles 21–23 to establish functional attachments with the epidermis. drl mutant animals show additional learning and memory defects as a consequence of structural defects of the adult brain (Bolwig et al. 1995; Dura et al. 1995; Moreau-Fauvarque et al. 1998; Simon et al. 1998; Grillenzoni et al. 2007; Sakurai et al. 2009).

RYK proteins interact with Wnt5, the cognate ligand for the transmembrane protein Fz. Wnt5 functions as a repulsive signal for Drl during embryogenesis, directing Drl-expressing axons away from the posterior commissure (PC) and toward the AC of each VNC hemisegment (Fradkin et al. 2004). Uncoordinated wnt5 mutants display axon navigation and axon fasciculation defects similar to drl mutant animals. Further, AC loss following wnt5 overexpression in the midline is dependent on functional Drl. Moreover, both drl and wnt5 mutants show alterations in dendritic branching of CNS serotonergic neurons (Singh et al. 2010). Like Drl, mammalian Wnt5a routes RYK expressing axons through the corpus callosum (Keeble et al. 2006) and corticospinal tract (Liu et al. 2005). Unlike mammal RYK (Lu et al. 2004), however, Drl does not impact on canonical Wnt/β-catenin signaling (Wouda et al. 2008) and the activity of Drosophila Wnt5 in PC repulsion appears independent of Fz (Yoshikawa et al. 2003). Interestingly, wnt5 itself appears to be a target of Drl-activated neurons (Fradkin et al. 2004; Yao et al. 2007).

Drl and Wnt5 positively regulate glutamatergic NMJ development and synaptic transmission. Both drl and wnt5 mutants have reduced numbers of synaptic boutons and reduced NMJ size (Liebl et al. 2008). Growth of the NMJ during larval stages requires coordination between the presynaptic motor neuron and postsynaptic muscle (Zito et al. 1999). drl mutant phenotypes at NMJs are rescued by muscle-specific expression of drl, whereas Wnt5 functions exclusively in presynaptic motor neurons. Reducing drl dosage suppresses NMJ overgrowth resulting from wnt5 neuronal overexpression, implicating Drl downstream from Wnt5. Further, rescue of drl phenotypes with exogenous Drl requires the WIF (Wnt-inhibitory-factor) domain. This evidence suggests that Wnt5 released from the presynaptic boutons binds Drl on postsynaptic muscle to regulate bouton growth and postsynaptic differentiation (Liebl et al. 2008).

Src64B and Src42A mutants show defects in mushroom body (MB) anatomy, salivary gland development, and AC formation similar to Wnt5 and drl mutants. Src64B interacts genetically with drl, as well as Wnt5, in MB development, salivary gland migration, and VNC neuron commissure formation (Nicola et al. 2003; Harris and Beckendorf 2007; Wouda et al. 2008). RYK proteins bear alterations in conserved catalytic residues required for phosphotransfer and are therefore inactive (Katso et al. 1999; Yoshikawa et al. 2001). Rather, Src64B functions as the TK responsible for Drl phosphorylation and interaction with Drl stimulates Src64B activity. Src64B is likely responsible for Wnt5/Drl-mediated axon repulsion, given that Drl-dependent axon pathfinding relies on Src64B kinase activity. Drl is proposed to provide substrate specificity for Src proteins (Wouda et al. 2008).

The learning and memory defects of drl mutant animals are likely a consequence of Wnt5 regulating olfactory circuitry. A subset of drl-expressing cells normally antagonizes Wnt5 produced by olfactory receptor neurons (ORNs) to appropriately pattern and orient glomeruli during antennal lobe development. This function of Drl relies on the WIF domain (Fradkin et al. 2004; Yao et al. 2007). Delineation of the functional requirement of drl in the antennal lobe refined cell-type identities to lateral neuroblast and ventral neuroblast-derived neurons (Bolwig et al. 1995; Dura et al. 1995; Moreau-Fauvarque et al. 1998; Simon et al. 1998; Grillenzoni et al. 2007; Sakurai et al. 2009). drl mutant phenotypes of the olfactory system resemble those resulting from wnt5 overexpression in ORNs: aberrant positioning of glomeruli and ectopic targeting of ORN axons to extraneous glomeruli structures. Moreover, enhancement or attenuation of Wnt5 function can either exacerbate or suppress drl phenotypes respectively (Fradkin et al. 2004; Yao et al. 2007).

A similar antagonistic relationship between drl and wnt5 exists in MBs. Pan-neuronal drl overexpression phenocopies wnt5 loss-of-function and attenuation of Drl activity can suppress wnt5 overexpression phenotypes. Neuronal drl expression is sufficient to nonautonomously rescue MB defects in drl mutants whereas wnt5 expression in MBs restores MB morphology to wnt5 mutants. These data suggest that Drl-dependent sequestration of Wnt5 is required to limit MB axonal growth (Bolwig et al. 1995; Dura et al. 1995; Moreau-Fauvarque et al. 1998; Simon et al. 1998; Grillenzoni et al. 2007; Sakurai et al. 2009).

Doughnut on 2: An RTK Involved in Migration

drl and doughnut on 2 (dnt) likely arose by gene duplication because the two genes display more similarity to each other than mammalian RYKs. Although Drl and Dnt share 60% identity, dnt cannot completely rescue drl mutant phenotypes (Oates et al. 1998). Maximal expression of dnt occurs 4–6 hr postfertilization, 2 hr before maximal drl expression; however, expression of both persists throughout the Drosophila life cycle (Roy et al. 2010). Like Drl, Dnt is considered catalytically inactive, based on substitutions of critical catalytic amino acids in the TK domain.

dnt is expressed initially in the central region and anterior domain of the embryo. Later expression occurs primarily in invaginating cells of the ventral furrow, gut, cephalic, and transverse furrow, and tracheal pits. The name doughnut comes from the rings of expression surrounding tracheal primordia (Oates et al. 1998; Savant-Bhonsale et al. 1999). dnt plays a minor role, with drl and Drl-2, in salivary gland cell migration during late embryogenesis (Harris and Beckendorf 2007). Disruption of the dnt locus influences multiple body size-related traits including face and head width, thorax length and wing size (Carreira et al. 2008). Although the disparate expression patterns of dnt and drl are indicative of paralogous function, to date relatively little characterization of dnt has substantiated this conjecture.

Derailed 2: An RTK Sharing Overlapping Roles with DRL in Olfactory Circuitry and Salivary Gland Migration

Derailed 2 (Drl-2) shares 35% identity with Drl, yet the two share distinct expression patterns and drl-2 mutants display relatively mild defects in antennal lobe development compared with drl: one of two displaced glomeruli displayed defects similar to wnt5 and drl mutants whereas the other was similar only to that of wnt5 and opposite to that of drl mutants. drl and drl-2 mutant alleles synergize in this context; a drl drl-2 double mutant displays additional defects resembling wnt5 mutants, implicating Drl-2 in Wnt5 signaling. Further, ORN-specific overexpression of wnt5 bears little effect in a drl-2 mutant or drl drl-2 double mutant background. Glial-specific expression of Drl-2 can compensate for loss of drl suggesting that these receptors have paralogous functions in Wnt5 signaling dependent on cell context (Bolwig et al. 1995; Dura et al. 1995; Moreau-Fauvarque et al. 1998; Simon et al. 1998; Grillenzoni et al. 2007; Sakurai et al. 2009).

Wnt5 in the CNS repels drl-expressing salivary gland tip cells, thereby dictating salivary gland migration. In this context, drl-2 mutants show similar defects to drl mutants: ventromedial curving of tip cells and a failure of visceral mesoderm attachment. drl drl-2 double mutant embryos phenocopy drl mutants, indicating that Drl-2 plays a minor role in salivary gland morphogenesis (Harris and Beckendorf 2007). Drl-2 has a similar role to Drl in the development of the larval and embryonic musculature; however, in this case drl-2 expression in specific motor neurons functions in preventing synapse formation with inappropriate ventral muscles (Inaki et al. 2007).

RET: THE HOMOLOG OF THE MAMMALIAN RTK PROTO-ONCOGENE RET

Ret expression begins ∼3.5 hr postfertilization in scattered regions throughout the yolk sac and is not detected again until 5–7 hr, in a subset of neuroblasts. At ∼10 hr, Ret is expressed in midline glia of the VNC, in the somatogastric nervous system anlage, in midgut precursor cells, and transiently in the malpighian tubule anlage. Late expression is observed in the developing PNS and CNS (Sugaya et al. 1994; Hahn and Bishop 2001; Fung et al. 2007). During larval stages, Ret is expressed in neuroendocrine cells of the brain and ventral ganglion, as well as in leg, wing, antennal, and eye imaginal discs (Hahn and Bishop 2001; Read et al. 2005). This expression pattern is similar to that of human RET, the closest vertebrate homolog of Drosophila Ret (52% identity in the TK domain).

The ligand for vertebrate RET, glial cell line-derived neurotrophic factor (GDNF), does not bind RET, but rather the GPI-linked coreceptor GFR-α. Although Drosophila Ret shares homology and structural organization in its extracellular domain with that of vertebrate RET, a homologous GDNF ligand or GPI-linked receptor does not exist in the Drosophila genome (Anders et al. 2001). Furthermore, Drosophila Ret does not bind GDNF. Rather, four extracellular cadherin-like domains suggest an ancient role for Ret in adhesion, although Drosophila Ret is incapable of self-association in vitro (Abrescia et al. 2005).

A Ret transgene with equivalent mutations to that observed in multiple endocrine neoplasia (MEN) expressed in the Drosophila eye caused phenotypes analogous to that in vertebrates: excessive proliferation and aberrant neuronal specification. In accordance with a proposed role for Ret in cell adhesion, deficiencies in the adhesion regulators Moe, Pax, and Cad-N2-enhanced Ret^MEN-dependent phenotypes (Hahn and Bishop 2001; Read et al. 2005). These phenotypes were further modulated by mutation of Ras, Src, and JNK, consistent with characterized roles for human RET (Arighi et al. 2005). Defective eye development induced by Ret^MEN expression was altered by DJ-1α/β—proteins linked to Parkinson’s disease—likely by modulating Ras/ERK signaling (Aron et al. 2010).

Cad96Ca: A CADHERIN DOMAIN-CONTAINING RTK INVOLVED IN WOUND REPAIR

Like Ret, Cad96Ca (also known as Stitcher) has both cadherin and TK domains (Tepass et al. 2000). Cad96Ca is expressed in all ectodermal epithelia during mid- and late embryonic stages but later becomes restricted to the epithelial optic lobe anlagen (Fung et al. 2007). Homozygous cad96C null animals die at late pupal stages. Cad96Ca displays TK activity in vitro, and a functional TK domain is required for rescue of Cad96Ca mutant animals (Wang et al. 2009).

Cad96Ca is predicted to play a role in nervous system development given that Cad96Ca expression was found down-regulated in late-stage embryos for which neuroectoderm was derived from glial cells (Egger et al. 2002). Consistent with this, many classical cadherins are expressed in the developing nervous system and have roles in neurite outgrowth, and axonal patterning and fasciculation (Tepass et al. 2000).

Cad96Ca is expressed in primordia of the spiracle—the external opening of the larval respiratory system—where it reinforces DE-Cad activity in posterior spiracle morphogenesis. Expression of Cad96Ca in this context depends of EGFR and Hh signaling, as well as the transcription factor Cut (Lovegrove et al. 2006; Maurel-Zaffran et al. 2010).

Cad96Ca facilitates embryonic re-epithelialization following wound healing by stimulating actin cable formation and the transcription of cuticle repair genes by the transcription factor Grh (Wang et al. 2009). Moreover, Cad96Ca can induce ERK phosphorylation, which is required for Grh activation (Mace et al. 2005).

Eph: AN RTK INVOLVED IN AXON PATHFINDING

Drosophila Eph displays similarity to both classes of vertebrate Eph receptors (∼35% in the extracellular region and 71% in the TK domain), whereas the best characterized ligand for Eph in Drosophila, Ephrin, shares ∼40% identity in the extracellular ephrin domain with both classes of human ephrin ligands (Bossing and Brand 2002).

Eph is expressed exclusively in the nervous system, initially in the neuroectoderm ∼5 hr postfertilization and then after ∼10 hr in a subset of neurons of the brain and VNC. Expression persists in the larval CNS and MBs, photoreceptor axonal projections and developing optic ganglia. Eph protein localizes to axons of elongating neurons, with highest concentrations in the growth cones of the earliest differentiating cortical and MB neurons and photoreceptors, and on longitudinal and commissural axons of the VNC (Scully et al. 1999; Dearborn et al. 2002; Boyle et al. 2006). This localization is proximal to that of Ephrin, which is concentrated in neuronal cell bodies along the outer edge of connectives and between commissures (Bossing and Brand 2002).

Disruption of Eph function by RNAi results in defective projection of photoreceptor axons as well as the aberrant targeting and loss of medulla and lobular cortical axons. Eph function is required specifically at the visual system midline to direct axon targeting in the developing eye and optic ganglia. Eph is predicted to fulfill a comparable role in the VNC; RNAi-mediated disruption of Eph, as for Ephrin, results in commissure fusion and loss, in addition to connective fragmentation as a consequence of interneuronal axon departure from the CNS longitudinal connectives. Ephrin expression at the midline, on the other hand, repels contralateral axon midline crossing and halts axonal growth along connectives, in an Eph-dependent manner. Eph null animals are viable and display abnormalities specifically in projection neuron targeting during MB development (Scully et al. 1999; Dearborn et al. 2002; Boyle et al. 2006). Additional phenotypes uncovered for Eph in other studies may reflect unintentional RNAi-mediated targeting of homologous targets.

Drosophila Vap33 is a proposed alternative ligand for Eph. Like Ephrin, Vap33 is membrane anchored and vap33 null mutants display MB defects in late pupae and adult brains identical to those of Eph. Moreover, inactivation of Eph can suppress muscular defects resulting from neuronal Vap33 expression. Vap33 can bind to the extracellular domain of C. elegans VAB-1 Eph receptor. This binding appears conserved among VAP proteins and is proposed to antagonize Ephrin binding (Tsuda et al. 2008).

Finally, Drosophila Ephexin (Exn) like its vertebrate Rho-type guanine nucleotide exchange factor counterpart binds Eph, at NMJs. Exn binds the Eph TK domain via its SH3 and Rho-GEF domains. Exn is required at the presynaptic nerve terminal of the NMJ to modulate synaptic vesicle release; disruption of either Exn or Eph interferes with homeostatic compensatory neurotransmitter release at NMJs. Eph is hypothesized to serve as a presynaptic receptor for a muscle-derived retrograde signal, speculated to be either Ephrin or Vap33 (Frank et al. 2009).

PDGF- AND VEGF-RECEPTOR RELATED: AN RTK WITH ROLES IN TISSUE SCULPTING, CELL MIGRATION, AND SURVIVAL

PDGF- and VEGF-receptor related (Pvr) is expressed ∼4 hr postfertilization in the procephalic mesoderm. Expression is later restricted to populations of scattered hemocytes, the hematopoietic cells in Drosophila. Three Pvr ligands, Pvf1-3, are expressed along stereotypical routes taken by migrating hemocytes. Although nonessential for differentiation, Pvr is required cell-autonomously for maintaining migrating populations of mature hemocytes. Pvr/Pvf induces ERK activation and hemocyte migration and is dependent on Ras1 (Heino et al. 2001; Cho et al. 2002; Brückner et al. 2004).

Plasmatocytes represent the majority class of hemocytes. They are phagocytic, clearing apoptotic debris generated during the programmed cell death that is necessary for tissue sculpting and metamorphosis (Tepass et al. 1994). Pvr mutant embryos show CNS axon scaffolding and glial cell positioning defects, as a consequence of reduced hemocyte numbers and therefore compromised neuron and glial cell elimination (Sears et al. 2003). They also show defective VNC condensation resulting from reduced hemocyte-derived extracellular matrix (Olofsson and Page 2005). Pvr is present in midline glia (MG) and all Pvf ligands localize to midline neurons of the CNS. Pvr/Pvf maintain and direct MG during embryogenesis; expression of activated Pvr at the midline induces enlargement of MG clusters and misallocated supernumerary MG, whereas ectopic Pvf expression at the midline or specifically in neurons reroutes MG migration. In the absence of functional Pvr, MG are disorganized or lost because of excessive apoptosis (Learte et al. 2008).

Embryonic plasmatocytes found larval hemocyte populations and self-renewal requires an intact PNS to attract plasmatocytes to a hematopoietic niche (Makhijani et al. 2011). Expression of activated Pvr stimulates larval hemocyte proliferation, whereas dominant negative Pvr has the opposite effect (Zettervall et al. 2004). At the onset of metamorphosis, lymph glands supply large numbers of plasmatocytes to phagocytose unnecessary larval tissue. Pvr is required in the lymph gland to regulate plasmatocyte differentiation and maintain levels of mature hemocytes (Jung et al. 2005).

Pvr is additionally expressed in ovarian border cells and is required for their initial migration in the direction of Pvf1, expressed by the developing oocyte (McDonald et al. 2003). Pvf1 engagement by Pvr provides directionality, because Pvr inactivation results in border cell clusters with misallocated and disoriented actin protrusions (Prasad and Montell 2007; Poukkula et al. 2011). The impetus for Pvr-directed border cell migration appears collectively to be the activation of the Rac-Mbc-ELMO complex (Duchek et al. 2001; Bianco et al. 2007; Wang et al. 2010), the accumulation of cortactin and cofilin at the migratory front of the cluster (Somogyi and Rørth 2004; Zhang et al. 2011) and the down-regulation of the transcriptional repressor Aop (Schober et al. 2005).

Pvr has been implicated in the migration of imaginal cells during metamorphosis. Pvr is required for JNK-dependent thorax closure. The Rac effector Crk-Mbc-ELMO complex links Pvr to JNK in this context, similar to border cell migration (Ishimaru et al. 2004). The rotation and dorsal closure of the male genital imaginal disc also relies on Pvr/Pvf1 to activate JNK (Macias et al. 2004). Mbc-ELMO functions downstream from Pvr additionally in epithelial cells stimulated to engulf their oncogenic neighbors. In this environment, JNK is the trigger for both apoptosis in mutant cells and Pvr activation in surrounding wild-type cells (Ohsawa et al. 2011).

Pvr mutants show defects in the anterior projection of renal tubules. Like CNS remodeling, these phenotypes derive from a lack of migrating hemocytes and a consequential lack of collagen secretion/deposition, which normally facilitates Dpp presentation by dorsal epidermal and visceral mesodermal cells and directs renal tubule migration. Pvf expression in the renal tubules attracts and activates Pvr-expressing hemocytes (Bunt et al. 2010).

Pvf1 and Pvf3 confine Pvr activity to the apical domain of the wing imaginal disc epithelium. Unrestricted Pvr activity results in a loss of epithelial polarity, ectopic adherens junctions, elevated basolateral actin filament polymerization, and neoplastic overgrowth (Rosin et al. 2004). Pvr and Pvf2 are expressed during heart metamorphosis, in cardiac valve precursors. Cardiac valves, dense accumulations of filamentous actin, are fewer following dominant-negative Pvr expression, whereas activated Pvr induces ectopic valve formation (Zeitouni et al. 2007). Congruent with a cardiac requirement for Pvr, reduction in Pvr function results in aberrant embryonic/larval heart pumping (Wu and Sato 2008). However, this phenotype may reflect defective neural circuitry associated with Pvr loss (Olofsson and Page 2005) rather than defective cardiac valve development (Zeitouni et al. 2007).

Like Ryk RTKs, Pvr is required autonomously to direct salivary gland migration toward the visceral mesoderm. Pvr is expressed in the developing salivary gland at the incipient site of placode invagination whereas Pvf1 is abundant at the tip of the migrating gland. Pvr mutants, like Pvf1 and Pvf2 mutants, have salivary glands that curve ventrally toward the CNS (Harris et al. 2007).

Pvf2 is expressed in the adult midgut in intestinal stem cells (ISCs) and enteroblasts. Expression increases with age and in response to oxidative stress, and correlates with age and stress-dependent increases in ISC populations. Ectopic Pvr expression specifically in ISCs and enteroblasts causes lethality. The guts of flies that survive display elevated numbers of proliferating cells that further amplify with age. Both Pvr and Pvf2 expression in ISCs stimulate their division and leads to altered differentiation—increased numbers of enteroendocrine cells at the expense of enterocytes (Choi et al. 2008). These Pvr-dependent effects rely on functional p38b ERK (Park et al. 2009).

Tie-LIKE RECEPTOR TYROSINE KINASE: ANOTHER RTK INVOLVED IN MIGRATION

Drosophila Tie-like RTK (Tie) shares ∼50% identity in its TK domain with the human Tie RTK (Ito et al. 1994). Tie expression peaks at late embryonic and pupal stages (Roy et al. 2010) primarily in the hindgut, salivary gland, and trachea (Chintapalli et al. 2007). Tie expression is up-regulated in Drosophila egg chambers: specifically in border cells and centripetal cells. This up-regulation is dependent on the basic region/leucine zipper transcription factor Slbo, a C/EBP homolog. Expression of dominant-negative Tie exacerbates border cell migration defects resulting from dominant-negative PVR and EGFR. This indicates that Tie has a redundant role with PVR and EGFR in directing border cell cluster migration (Wang et al. 2006).

Tie likely plays a role in development of the Drosophila sensory system. Tie expression is up-regulated in Drosophila imaginal discs on overexpression of the Pax6 homolog Eyeless, a transcription factor directing neuronal differentiation in the retina. Tie is a predicted target of Eyeless, based on in silico approaches. Tie is expressed in the eye imaginal disc and is predicted to function early in retinal differentiation (Michaut et al. 2003; Ostrin et al. 2006).

CONCLUDING REMARKS

The breadth of scenarios wherein RTK signaling contributes to Drosophila development is undeniably impressive. RTKs provide a means of communication between different tissues and cell types that contributes to a robust and highly reproducible developmental program. RTKs vary dramatically in their expression with respect to cell type, ranging from ubiquitous expression, as in the case of InR for instance, to restricted expression in specific subsets of cells, as for Btl. Most interestingly, the activity of a few RTKs is exploited many times over in different developmental contexts, such as for EGFR, although the activity of other RTKs, for instance Torso and Sev, is highly specific to a few specialized functions. Remarkably, many downstream effectors are shared among different RTKs (Table 1) and their hierarchical organization is reiterated in various biological contexts. This implies that specificity in signaling output is likely rendered by a limited number of factors. For example, the insulin receptor substrate Chico is specific for InR signaling (Bohni et al. 1999), whereas the adaptor protein Shc functions downstream from Torso, EGFR, and Btl, but not Sev (Luschnig et al. 2000; Cabernard and Affolter 2005). Additionally, the context and manner in which ligands are presented to the RTK is likely to influence outcome. For instance, the overexpression of pyr and ths in combination elicits a phenotype opposite to that of individual overexpression of pyr or ths (Kadam et al. 2012). Furthermore, heparan sulfate proteoglycans are required for delivery and stabilization of ligands with RTKs and therefore maximal activation of signaling (Nybakken and Perrimon 2001). Moreover, the competency and direction by which a cell to responds to RTK activation relies in large part on the transcription factors expressed in that particular cell type. Because of its limited genetic redundancy, Drosophila remains an attractive model in terms of RTK pathway component identification, and has successfully served to uncover corresponding conserved vertebrate counterparts. Components rendering fine-tuning functions in many well-characterized RTK pathways continue to be discovered. Astonishingly, relatively little is known regarding downstream signaling evoked by approximately half of all Drosophila RTKs. The current availability of genome-wide transgenic RNAi reagents in Drosophila will undoubtedly expedite further investigation of these RTKs to alleviate this knowledge gap.