Journ

alof

Cell

Scie

nce

The Drosophila Arf GEF Steppke controls MAPKactivation in EGFR signaling

Ines Hahn, Bernhard Fuss, Annika Peters, Tamara Werner, Andrea Sieberg, Dominic Gosejacob andMichael Hoch*LIMES-Institute, Program Unit Development, Genetics and Molecular Physiology, Molecular Developmental Biology, University of Bonn,Carl-Troll-Str. 31, D-53115 Bonn, Germany

*Author for correspondence (m.hoch@uni-bonn.de)

Accepted 7 March 2013Journal of Cell Science 126, 2470–2479� 2013. Published by The Company of Biologists Ltddoi: 10.1242/jcs.120964

SummaryGuanine nucleotide exchange factors (GEFs) of the cytohesin protein family are regulators of GDP/GTP exchange for members of theADP ribosylation factor (Arf) of small GTPases. They have been identified as modulators of various receptor tyrosine kinase signalingpathways including the insulin, the vascular epidermal growth factor (VEGF) and the epidermal growth factor (EGF) pathways. These

pathways control many cellular functions, including cell proliferation and differentiation, and their misregulation is often associated withcancerogenesis. In vivo studies on cytohesins using genetic loss of function alleles are lacking, however, since knockout mouse modelsare not available yet. We have recently identified mutants for the single cytohesin Steppke (Step) in Drosophila and we coulddemonstrate an essential role of Step in the insulin signaling cascade. In the present study, we provide in vivo evidence for a role of Step

in EGFR signaling during wing and eye development. By analyzing step mutants, transgenic RNA interference (RNAi) andoverexpression lines for tissue specific as well as clonal analysis, we found that Step acts downstream of the EGFR and is required forthe activation of mitogen-activated protein kinase (MAPK) and the induction of EGFR target genes. We further demonstrate that step

transcription is induced by EGFR signaling whereas it is negatively regulated by insulin signaling. Furthermore, genetic studies andbiochemical analysis show that Step interacts with the Connector Enhancer of KSR (CNK). We propose that Step may be part of a largersignaling scaffold coordinating receptor tyrosine kinase-dependent MAPK activation.

Key words: Arf GEF Steppke, Drosophila, EGFR signalling, MAP kinase, RAS

IntroductionThe proper development of multicellular organisms requires the

coordination of proliferation and differentiation, which is a

particular challenge during the formation of the tissues and

organs of the body. Numerous studies have shown that receptor

tyrosine kinases such as the vascular growth factor receptor

(VEGFR; Mannell et al., 2012) epidermal growth factor receptor

(EGFR) and insulin/insulin-like growth factor receptors (InR/

IGF-Rs) play prominent roles in signaling cell proliferation and

differentiation (reviewed by Schlessinger, 2004). Misregulation

of both pathways is often causative for tumor development and

progression through their effects on uncontrolled cell growth,

inhibition of apoptosis, angiogenesis, and tumor-associated

inflammation (Witte et al., 2009; Alvarez et al., 2010;

Antonarakis et al., 2010; Arkenau, 2009). Determining how

growth and differentiation are coordinated by these pathways is

thus essential to understanding normal development, as well as

disease states such as cancer.

We have recently identified Steppke (Step) as a new and

essential component of the insulin signaling pathway in Drosophila

(Fuss et al., 2006). The insulin signaling cascade is conserved from

flies to humans and was shown to regulate cell and organismal

growth in response to extrinsic signals such as growth factors and

nutrient availability. In Drosophila, activation of a unique insulin-

like receptor (InR) stimulates a conserved downstream cascade

that includes the Phosphatidylinositol-3-kinase (PI3K) and

Protein Kinase B (PKB or AKT). AKT is involved in

enhancement of glucose absorption and glycogen synthesis,

and regulates the activity of the Forkhead box O (FoxO)

transcription factor, a negative regulator of cell growth

(Teleman, 2010; Junger et al., 2003). We could show that

Step acts downstream of the insulin receptor and upstream of

PI3K in the insulin/IGF-like signaling (IIS) cascade (Fuss et al.,

2006). Step is a member of the cytohesin family of guanine

nucleotide exchange factors (GEFs) which regulate small

GTPases of the ADP-ribosylation factor (ARF) family

(Kolanus, 2007). Small ARF GTPases are involved in the

regulation of many cellular processes including vesicle

transport, cell adhesion and migration. Studies in mice have

confirmed an evolutionary conserved role of cytohesin family

members in IIS (Hafner et al., 2006).

Whereas our previous studies focused on the role of Step in

IIS-dependent larval growth control, we now examined its

function in the Drosophila wing, which develops from an

epithelial sheet during larval and pupal stages (Bier, 2000; De

Celis, 2003). The wing is an ectodermal structure formed by a

dorsal and ventral epithelium, interspersed with cuticular

ectodermal tubes, the so called wing veins. Stereotypical

arrangement of wing veins is determined in the imaginal wing

disc in late larval and pupal stages by several signaling pathways

including the EGFR cascade (Martın-Blanco et al., 1999). EGFR

activation by EGF-like ligands Spitz or Vein results in the

2470 Research Article

Journ

alof

Cell

Scie

nce

activation of the small GTPase RAS by its loading with

guanosine triphosphate (GTP), which as a result triggers the

activation of a number of downstream effector proteins including

the Ser/Thr-kinase RAF [mitogen-activated protein kinase

(MAPK) kinase kinase]. Once activated, RAF phosphorylates

and activates MEK (MAPK kinase), which in turn phosphorylates

and activates MAPK/ERK (Cobb and Goldsmith, 2000).

Phosphorylated MAPK exerts its role in the cytoplasm as well

as in the nucleus, where it controls expression of EGFR target

genes like pointed (pnt), argos (aos), rhomboid (rho) and ventral

nervous system defective (vnd) (Klambt, 1993; Gabay et al.,

1996; Wasserman and Freeman, 1998). The scaffolding protein

connector enhancer of KSR (CNK) has been described to

facilitate RAS/RAF/MAPK signaling by providing a protein

scaffold at the plasma membrane that integrates Src and RAS

activities to enhance RAF and MAPK activation (Claperon

and Therrien, 2007). EGFR/MAPK signaling is crucial quite

early during wing vein differentiation, where phosphorylation of

MAPK determines the positioning of proveins and later during

development for maintenance of longitudinal veins (Blair, 2007).

In addition to patterning, both EGFR/RAS/MAPK signaling and

IIS control general cell proliferation and cell growth during wing

development. Thus, EGFR/RAS/MAPK signaling controls both

cell fate (vein versus intervein) and general cell proliferation

along with IIS at similar times within the wing tissue.

Recent studies in human lung and breast adenosarcoma

cancer cell lines indicated a function of cytohesins in ErbB

(EGFR) signaling, where they facilitate signaling by stabilizing

an asymmetric ErbB receptor dimer (Bill et al., 2010). Here

we provide the first in vivo model that the cytohesin Step,

in addition to its previously characterized function as component

of IIS, regulates EGFR signaling dependent wing growth and

vein differentiation. Our genetic, immunohistochemical and

biochemical experiments indicate that Step acts downstream of

the EGF receptor in the EGFR signaling cascade and is

necessary and sufficient for MAPK activation and the

induction of EGFR target genes. Whereas step transcription is

negatively regulated by IIS (Fuss et al., 2006), it is induced by

EGFR signaling. We further provide evidences that Step might

directly interact with the Connector Enhancer of KSR (CNK)

protein that is part of a protein scaffold known to coordinate

RAS-dependent RAF and MAPK signaling from tyrosine kinase

receptors.

ResultsThe Step protein contains three characteristic motifs, a Sec7

domain responsible for the GEF activity, a coiled-coil domain

(CC) mediating protein-protein interactions and a pleckstrin

homology (PH) domain (Fig. 1A). The PH domain is required

for plasma membrane recruitment via specific binding to

phosphatidylinositol-3,4,5-trisphosphate, the second messenger

generated by class I PI3Ks (Carnero, 2010). We previously

studied the function of Step mainly during larval development

where it is acting as a component of the IIS cascade controlling cell

and organismal size (Fuss et al., 2006; Becker et al., 2010).

Expression studies using an anti-Step antibody (see Materials and

Methods) indicate, however, that step is also expressed in imaginal

discs and during pupal and adult stages. In wing imaginal discs we

found ubiquitous expression of step in a vesicle-like pattern with

an accumulation of the protein at the cell cortex (Fig. 1B9)

consistent with previously described localization of the PH domain

(Britton et al., 2002).

Step is required for wing vein differentiation

Homozygous stepK mutants are lethal at pupal stages (Fuss et al.,

2006). Expression of UAS-step rescues hypomorphic stepK

mutants to adulthood. Whereas this complete rescue requires a

strong induction of step (Fig. 2C) a weak expression of step in

homozygous stepK mutants leads to only a partial rescue

including defects in wing vein differentiation (Fig. 2B).

Fig. 1. Step is expressed ubiquitously in imaginal wing discs. (A)

The cytohesin Steppke (Step) harbours three major domains: a

catalytically active Sec7 domain (Sec7), a C-terminal pleckstrin

homology (PH) domain and an N-terminal coiled-coiled (CC) domain.

Protein fragment used for immunization (dashed line) and target

sequence of stepRNAi are indicated. Three UAS constructs were used in

this study: a full length step version, a full length (FL) step construct

with an N-terminal GFP tag and a variant lacking the PH domain. (B–

B0) Third instar imaginal discs stained for E-Cadherin (B) and Step

(B9,B0). Step is expressed ubiquitously in imaginal wing discs and

localizes in a vesicular pattern (see inset in B9 and B0) in close

proximity to the cell cortex (marked by E-Cadherin). (C,D) Third instar

imaginal wing discs stained for Step. Anterior is right. (C–C0) Step

antibody detects ectopic step expression in the posterior imaginal wing

disc. Please note that detection channel intensities are adjusted to avoid

overexposure. (D–D0) Expression of stepRNAi via enGal4 leads to

reduced Step protein levels in the posterior wing disc.

Steppke controls EGFR/MAPK signaling 2471

Journ

alof

Cell

Scie

nce

To complement the genetic analysis of step gene function (see

Materials and Methods) based on hypomorphic step alleles and a

step deficiency described previously (Fuss et al., 2006; Becker

et al., 2010) we generated step RNA interference (RNAi) and

overexpression lines enabling tissue specific as well as clonal step

knockdown and gain of function experiments. As a driver line, we

used engrailed-Gal4 (enGal4) which mediates expression in the

posterior compartment of the wing, the anterior compartment

serving as internal control (Tabata et al., 1995).

UAS-stepRNAi-mediated knockdown of step in the posterior

wing compartment results in a strong reduction of Step protein

levels (Fig. 1D–D0) and causes a reduction of the wing size

Fig. 2. Step is required for wing vein differentiation. (A) Wild-type adult Drosophila wings contain five lateral wing veins (L1–L5) and two cross veins

(anterior cross vein, acv; posterior cross vein, pcv). Yellow pseudocolouring indicates the Engrailed expression domain (Simpson, 2007). Anterior is up, distal is

right. (B,C) Expression of step via hsFLP rescues wing vein defects of stepK/stepK mutants. Depending on the duration of the heat shock, rescue is either

partial (B) or complete (C). (D–K) enGal4 was used to express gene specific RNAi constructs or gain of function transgenes respectively in the posterior wing

compartment (see yellow area in A) and to score for wing vein differentiation and growth in mutant clones (posterior compartment) versus unaffected control

clones (anterior compartment). stepRNAi expression (D) in a wild-type background results in loss of the cross veins (arrows); in a haploinsufficient step deficiency

background parts of L4 (asterisks) are lost as well (F). Full length step (E) but not stepDPH overexpression (G) leads to excess wing vein tissue (see arrows).

Reduction of IIS by InRRNAi (H) leads to a decreased posterior wing area (see M for quantification); but wing vein tissue is not lost. Enhanced IIS by InR

overexpression (I) results in increased wing area posterior from lateral vein L4 (for quantification see M); wing vein formation is not affected. A reduction of

EGFR signaling via expression of EGFRRNAi results in loss of wing veins (see arrows in J). acv and L4 are lost almost entirely. EGFR signaling gain of

function by EGFR overexpression leads to additional wing vein tissue in the posterior wing (see asterisks in K). (L,M) The effect of step modulations on wing size

was assessed by measuring the area of the posterior part of the wing (see yellow area in inset) relative to the total wing area. Yellow bar indicates size range

of control wings. (L) step loss (stepRNAi) and gain of function (UAS-step) leads to an decrease in wing size; ectopic expression of a step variant lacking the

pleckstrin homology (PH) domain (stepDPH) does not affect wing size. (M) Activation of IIS by InR or dp110 expression results in significantly larger wings,

whereas inhibition of IIS by InRRNAi decreases wing size significantly. Activation and inhibition of Egfr signaling (Egfr and EgfrRNAi expression) resulted in

significantly smaller wing areas. n.20. Error bars indicate s.e.m. Statistical significance was analyzed by Student’s t-test. P-values are indicated.

Journal of Cell Science 126 (11)2472

Journ

alof

Cell

Scie

nce

(Fig. 2L) due to a reduction in cell number but not size(supplementary material Fig. S1). Furthermore, loss of wing

vein tissue can be observed and both cross veins are not formedproperly (Fig. 2D, compare with wild-type wing in Fig. 2A).Reducing step levels further by RNAi mediated knockdown in

the background of a step heterozygous deficiency (stepdef,Materials and Methods), the phenotype is enhanced and lateralvein L4 displays gaps in addition (Fig. 2F). In contrast,overexpression of step in the posterior wing compartment

induces the formation of ectopic longitudinal and cross veintissue and excess veins at the distal end of lateral vein L4(Fig. 2E). Of note, overexpression of a step variant lacking the

PH domain (stepDPH) does not result in excess wing vein tissue(Fig. 2G), indicating that membrane recruitment of Step isessential for its function on vein differentiation.

The wing growth phenotypes observed upon modulation ofstep levels are similar to the effects observed when InR is reduced(Fig. 2F,H,L,M) (Bohni et al., 1999; Brogiolo et al., 2001),

consistent with an expected role of Step in IIS. However, wingvein defects or ectopic veins which are found upon step reductionor overexpression, respectively, cannot be explained by a role in

IIS, since vein formation is not affected in wings expressingUAS-InR or UAS-InRRNAi (Fig. 2H,I). Indeed, they ratherresemble EGFR-dependent phenotypes in the wing. ReducingEGFR levels by RNAi-mediated knockdown in the posterior

compartment leads to loss of lateral vein L4 and the apical crossvein (Fig. 2J) (Martın-Blanco et al., 1999) and growth defects(Fig. 2M). Since expression of several transgenes activating

EGFR signaling (like UAS-RasV12, UAS-Rafgof or the activatedform of the EGFR) via enGal4 lead to early larval lethality, wemade use of an UAS-EGFR constructs to mildly induce the EGFR

pathway. Overexpression of the wild-type receptor results in mildEGFR signaling gain of function effects: excess wing vein tissueis formed between L4 and L5 (Fig. 2K). This effect is more

severe but similar to phenotypes observed for Step modulation.

Step induces EGFR target genes and is required for MAPKactivation

To further investigate a potential role of step in EGFR signaling,we analyzed the effect of step on the EGFR target genes argos

(aos), pointedP2 (pnt), and rhomboid (rho). All of these targetgenes were significantly reduced in hypomorphic stepK mutants(Fig. 3A) as well as wing discs expressing stepRNAi (Fig. 3B).

EGFR target genes are induced in imaginal eye discsoverexpressing step (Fig. 3C). To further study the impact ofstep on EGFR target gene expression, we analyzed Aos protein

levels in imaginal wing discs and found enhanced Aos levels inclones expressing step (Fig. 3D–D0). Similarly, overexpressionof step in the posterior compartment of an imaginal wing disccarrying the reporter aos-lacZ induces ectopic Argos expression

in the posterior wing disc (Fig. 3E–E0).

To test where Step may act in the EGFR cascade we analyzed

MAPK activity in imaginal wing discs, in which step expressionwas reduced or elevated in the posterior wing compartment.

It is known that the EGFR signals are transduced by a cascade

that includes RAS, RAF and MEK, which dually phosphorylatesMAPK directly to generate phospho-MAPK (pMAPK) (reviewedby Wasserman and Freeman, 1997). pMAPK then disassociates

from MEK and forms active homodimers that can phosphorylatetargets in the cytoplasm, and/or translocate to the nucleus toactivate nuclear targets, and thus controls gene expression (Brunet

et al., 1999; Chen et al., 1992; Lenormand et al., 1998). To monitor

MAPK activity, we made use of an antibody specific for thedoubly phosphorylated form of MAPK (pMAPK).

Reduction of step levels causes an inhibition of MAPKphosphorylation (Fig. 3F–F0) whereas overexpression of step in

the posterior wing compartment results in elevated pMAPKlevels (Fig. 3G–G0). Induction of EGFR signaling by EGFR

overexpression in imaginal wing discs results in similarly

increased pMAPK levels (supplementary material Fig. S2D–D0). Of note, expression of the InR does not lead to increasedpMAPK levels (supplementary material Fig. S2F–F0), indicating

that MAPK activation does not occur in response to a Stepdependent IIS function.

Step acts downstream of the EGF receptor in the EGFRpathway

To identify the level at which step is involved in with the EGFRpathway we tested whether step and several EGFR pathwaycomponents interact genetically, similar to approaches published

recently (Gaengel and Mlodzik, 2003; Yoshida et al., 2004).Knocking down step transcript and protein levels in an EGFRRNAi

background using enGal4 increases the effect of EGFRRNAi in

the posterior compartment of the developing wing (compareFig. 4A,A9). Lateral vein L4 was entirely lost, both cross veinswere not differentiated and the distal end of lateral vein L5 was

missing (see arrows in Fig. 4A). Similarly, stepRNAi enhanceswing vein defects caused by reduction of RAS, RAF and MAPKtranscript levels (compare Fig. 4B,B9, Fig. 4C,C9 and

Fig. 4D,D9). Furthermore, UAS-step expression could rescuewing vein defects caused by EGFRRNAi (Fig. 4A0) as well as itimproves wing vein phenotypes due to RASRNAi expression(Fig. 4B0). In contrast, wing vein differentiation defects caused

by RAFRNAi or MAPKRNAi were not rescued upon UAS-step

expression (Fig. 4C0,D0). There results point to a function of Stepbetween RAS and RAF in the EGFR pathway.

Similar to the genetic interaction of step and EGFR pathwaycomponents observed during wing vein development, we foundstep to genetically interact with the EGFR component Star inanother EGFR dependent process. An analysis of eye development

in late stepk/k pupae clearly shows a loss of photoreceptors inthe step hypomorph (supplementary material Fig. S3A9,A0)which is also found in EGFR mutants (but not in mutants for the

insulin receptor; McNeill et al., 2008). Flies heterozygote for theStarIIN allele display a very mild loss of photoreceptorsdifferentiation phenotype (supplementary material Fig. S3A9)

which can be enhanced by removal of one copy of essential EGFR/MAPK signaling components (Gaengel and Mlodzik, 2003). Asimilar phenotype enhancement occurs in flies transheterozygousfor the stepK and the StarIIN allele: the mild defects of

photoreceptor differentiation observed in heterozygous SIIN

mutants (supplementary material Fig. S3B) are significantlyincreased in transheterozygosity with stepK mutants (SIIN/+;

stepK/+; supplementary material Fig. S3C). The alleliccombination SIIN/+; InR05545/+ does not lead to an enhancedphenotype, indicating an IIS independent function of step during

photoreceptor development. Furthermore, Step function alsomediates MAPK activation in context of eye development. Aclonal reduction of step activity in imaginal eye discs via UAS-

stepRNAi results in lowered pMAPK levels (supplementary materialFig. S3C,C9), whereas clonal overexpression of step leads toincreased pMAPK levels (supplementary material Fig. S3D,D9).

Steppke controls EGFR/MAPK signaling 2473

Journ

alof

Cell

Scie

nce

The observations on the role of Step in EGFR signaling during

wing and eye development are consistent with reports obtained in

mammalian cell culture models where cytohesins have been

implicated with ErbB (EGFR) in human lung and breast

adenosarcoma cancer cell lines. Bill et al. could show that

cytohesin 2 is a cytoplasmic ErbB receptor activator by

stabilizing active asymmetric ErbB dimers (Bill et al., 2010).

Interestingly, our data point towards a more downstream role of

Step in the signaling cascade, since step overexpression is able to

rescue EGFRRNAi and RasRNAi wings. Moreover, our data point to

a function of Step upstream of MAPK in the EGFR pathway.

Step interacts with the scaffolding protein dCNK to control

EGFR/MAPK signaling

The Connector Enhancer of KSR (CNK) protein family has been

proposed to function as protein scaffolds (Therrien et al., 1998;

Douziech et al., 2003). Data obtained in HepG2 cell experiments

indicate that CNK1 binds the coiled-coil domain of cytohesin 2

via a C-terminal region (CBD, cytohesin binding domain) and

mediates plasma membrane association via its PH domain. At the

plasma membrane, cytohesin 2 activates ARF6 and thereby

enhances PIP5K activation. This leads to PIP2 enrichment at the

plasma membrane, which in turn promotes IRS membrane

recruitment and thereby facilitates IIS (Lim et al., 2010). In

contrast, the Drosophila CNK homologue, dCNK, has been

associated with EGFR/MAPK signaling: dCNK directly

associates with the kinase domain of RAF via a short amino-

acid sequence, called the RAF-interacting motif (RIM), and

modulates RAF activity according to the RTK signaling status

(Douziech et al., 2006; Laberge et al., 2005). Without RTK

signals, CNK-bound RAF is inhibited by a second motif adjacent

to the RIM, called the inhibitory sequence (IS). Upon RTK

activation, CNK integrates Src and RAS activities, which then

leads to RAF activation.

Fig. 3. Step can induce EGFR/

MAPK signaling. enGal4 was used to

express gene specific RNAi constructs

or gain of function transgenes

respectively in the posterior wing

compartment. EGFR target gene

expression in homozygous stepK larvae

(A) as well as in imaginal wing discs of

third instar (L3) larvae expressing

UAS-stepRNAi (B) or UAS-step (C). step

knockdown in stepK mutants and UAS-

stepRNAi expressing imaginal wing

discs as well as UAS-step

overexpression was validated. n$5,

error bars indicate s.e.m. Statistical

significance was analyzed by Mann–

Whitney U-test. P values are as

indicated: *P,0.05; **P,0.01; aos,

argos; pnt, pointed; rho, rhomboid;

step, steppke. (D,E) Posterior imaginal

wing discs overexpressing UAS-step

either clonally via hsFLP (D9,D0) or in

the posterior compartment via enGal4

(E9,E0) show elevated Aos protein

levels (D) or b-Gal reporter levels in

the aos-lacZ line (E), respectively.

(F,G) Imaginal wing of third instar

larvae stained for doubly

phosphorylated MAPK (pMAPK).

Posterior expression of transgenes in

the imaginal wing disc compartment

via enGal4 is marked by coexpression

of GFP. (F9,F0) step knockdown results

in less phosphorylation of MAPK in

this area (see arrow) compared with the

anterior compartment (F). Increased

pMAPK levels are seen in step

overexpressing posterior wing discs

(see asterisk in G) compared with

unaffected control compartment (no

GFP signal in G9 and G0).

Journal of Cell Science 126 (11)2474

Journ

alof

Cell

Scie

nce

To test whether Step function in EGFR signaling during wing

differentiation is at the level of dCNK we tested genetic and

biochemical interaction of the two factors. We monitored CNK

expression using a CNK antibody (Douziech et al., 2003). dCNK

is expressed ubiquitously in the wing imaginal disc in a vesicle-

like pattern (supplementary material Fig. S4A). dCNK

knockdown in the posterior compartment of the imaginal wing

disc using enGal4 in combination with UAS-dCNKRNAi results in

loss of wing vein tissue (Fig. 5A; supplementary material Fig.

S4A9), consistent with its described role in EGFR/MAPK

signaling. This reduction of vein tissue can be further increased

when step is downregulated in addition (Fig. 5A9). In contrast,

step overexpression in a dCNKRNAi background rescues the

wing vein differentiation defects (Fig. 5A0) indicating that

Fig. 4. Step acts downstream of the EGF receptor in the EGFR pathway. (A–D) Knockdown of the EGFR pathway components EGFR, RAS, RAF and

MAPK in the posterior wing compartment via RNAi leads to reduction of wing vein differentiation (see arrows in A9–D9). For each EGFR pathway component

knockdown, coexpression of stepRNAi enhances the loss of wing vein differentiation (compare A and A9, B and B9, C and C9, D and D9). UAS-step expression can

rescue wing differentiation defects in UAS-EGFRRNAi wings (A0) as well as enhance wing vein differentiation defects caused by RASRNAi (B0). In contrast,

wing vein defects in RAFRNAi and MAPKRNAi wings were not rescued by UAS-step (compare C9 and C0, D9 and D0).

Fig. 5. Step interacts with CNK. Knockdown of CNK levels in the

posterior wing leads to loss of wing veins (A). The CNKRNAi phenotype can

be increased by coexpression of stepRNAi (compare A and A9). step

overexpression is able to rescue wing vein differentiation defects caused by

CNK knockdown (A0). (B) Direct interaction of Step and CNK was tested

in a yeast two-hybrid approach. The CNK C-terminus (aa 818–1557) was

fused to the Gal4 activation domain of plasmid pGAD. Binding of three

Step deletion variants fused to the Gal4 DNA-binding domain of plasmid

pGB was analysed: Versions lacking the coiled-coiled (CC; stepDCC), the

Sec7 (stepDSec7) and the PH domain (PH; stepDPH) were cloned. Growth in

the two-hybrid assay is indicated by ‘+’; ‘2’ indicates the inability of

corresponding transformants to grow on selection medium. Step mutants

lacking the PH and the Sec7 domain were tested positive in this binary

test system.

Steppke controls EGFR/MAPK signaling 2475

Journ

alof

Cell

Scie

nce

Fig. 6. Step levels are increased upon EGFR overexpression. Clonal expression of EGFR and GFP (A,B) or GFP only (C) in third instar imaginal wing

discs stained for EGFR, Step and Schlank (as a negative control). Clonal EGFR overexpression via hsFLP leads to increased Step levels (A0), whereas Schlank

levels are not affected (B9). Clonal GFP expression alone does not influence Step levels (C–C0). step expression is induced in L3 larvae overexpressing UAS-

EGFR (D; induction of EGFR signaling validated by aos induction) and downregulated in specimens expressing UAS-EGFRDN (E; inhibition of EGFR

signaling validated by downregulation of aos). Identically treated L3 larvae of the genotype hsGal4;+ served as control. n$5; error bars indicate s.e.m. Statistical

significance was analyzed by Mann–Whitney U-test. P values are as indicated. aos, argos; step, steppke. (F) Model of potential function of Step in wing

vein and eye development. Step acts in the EGFR pathway downstream of CNK. Step expression is regulated by EGFR signaling.

Journal of Cell Science 126 (11)2476

Journ

alof

Cell

Scie

nce

Step acts downstream of dCNK in the EGFR/MAPK cascade.

Both Step and dCNK are localized in a vesicle-like patternthroughout the cells. Subcellular localization studies reveal apartial co-localization of Step and dCNK in the cortical region

of cells (supplementary material Fig. S4C–C0; forquantification see supplementary material Fig. S4C9,C0). Toaddress whether Step has the potential to directly interact withdCNK we used the binary yeast-two-hybrid system (see

Materials and Methods). We tested interactions of Stepprotein variants that were lacking either one of the threefunctional domains of Step, the coiled-coil, the Sec7 or the PH

domain, respectively, with the C-terminus of dCNK (seeFig. 5B; supplementary material Fig. S4B). The resultssuggest that Step, similar to data obtained in mammalian cell

culture models, might interact with the C-terminus of CNK viaa stretch containing the coiled-coil domain and thereby mightact as part of the CNK scaffolding complex.

EGFR signaling induces Step expression

The importance of Step for EGFR signaling may further beindicated by the following observation: upon induction ofthe EGFR pathway, Step levels are enhanced. Clones in

imaginal wing discs overexpressing UAS-EGFR show increasedlevels of Step protein (Fig. 6A–A0). This increase cannot be seenwhen only GFP is clonally expressed (Fig. 6C–C0). The

modulation of Step levels upon EGFR pathway activity canfurther be seen on transcript levels. Overexpression of UAS-

EGFR leads to a strong induction of step transcript levels

(Fig. 6D), whereas overexpression of a dominant negativeversion of the EGFR inhibiting the signaling pathway leads toa downregulation of step (Fig. 6E). This indicates that step

transcriptional levels are positively regulated by the EGFRpathway as a part of a positive feedback loop. This is opposite toeffects seen on step levels upon IIS modulation, where step

transcription is negatively regulated by the pathway and activated

by FOXO (Fuss et al., 2006).

DiscussionOur study demonstrates an in vivo function of the Arf GEF Step

as an essential component of the EGFR signaling pathwaywhich acts downstream of the EGFR (Fig. 6F). Step is necessaryand sufficient for activation of MAPK and the induction of

EGFR target genes in the Drosophila wing. Based on ourbiochemical, immunohistochemical and the genetic data wepropose a mechanistic model in which Step and dCNK

interaction is important for EGFR signaling. dCNK is thesingle member of the CNK protein family in Drosophila. CNKproteins are scaffolding proteins that have been linked withRAS, Rho, Rac, Ral and Arf GTPases and are proposed to act as

general regulators of GTPase-mediated events downstream ofreceptor tyrosine kinases, including EGFR and InR/insulin-likegrowth factor receptors (Claperon and Therrien, 2007). Together

with the kinase suppressor of RAS (KSR), CNK was shown toassemble a signaling complex including RAF and MEK whichpromotes RAS-dependent RAF activation and the subsequent

phosphorylation of MAPK (Roy et al., 2002; Therrien et al.,1998; Douziech et al., 2006; Rajakulendran et al., 2008). Wesuggest that Step is a functional part of this scaffolding complex

via its direct interaction with CNK. This is also consistent withrecent data in HeLa and 393T cells showing that human CNK1directly interacts with cytohesin-2 to coordinate PI3K/AKT

signaling downstream of InR/IGF-R (Lim et al., 2010). It was

proposed that CNK1 recruits cytohesin-2 to the plasmamembrane, where activity of plasma membrane boundGTPases leads to a PIP2 rich microenvironment, which

enhances IRS1 recruitment and hence facilitates PI3K/AKTsignaling (Lim et al., 2010). Similarly, Drosophila cytohesinStep was shown to be required for PI3K activation (Fuss et al.,2006). Together, several lines of evidence support a role of

cytohesins and CNK in similar signaling contexts (RAS/RAF/MAPK and PI3K/AKT signaling), where a direct interaction ofboth proteins as part of a signaling platform might promote

downstream signaling events like MAPK phosphorylation andPI3K activation. This does not exclude other functions ofcytohesins, e.g. the stabilization of asymmetric ErbB (EGFR)

dimers, as shown recently in human lung and breastadenosarcoma cancer cell lines (Bill et al., 2010). Our dataindicate, however, that a major function of the Drosophila

cytohesin Step in EGFR signaling resides downstream of theEGFR and upstream of MAPK.

Materials and MethodsFly stocks

Flies were raised on standard fly food at 25 C if not mentioned otherwise. Thefollowing fly stocks were used: wild-type (Oregon R); stepK08110 (Bloomingtonstock #10770); Df(2L)TW65 (referred as stepdef, Bloomington stock #1602);enGal4, UAS-Dcr2, UAS-GFP (Bloomington stock #25752); hsFLP;act .CD2

.Gal4, UAS-GFP (Bloomington stock #9431); UAS-EGFR (a gift of C. Klambt);UAS-dp110CAAX (Bloomington stock #25908); UAS-step (Fuss et al, 2006); UAS-InRA1325D (Bloomington stock #8263).

The following stocks were obtained from the Vienna Drosophila RNAi Centre(Dietzl et al., 2007): UAS-EGFRRNAi (VDRC stock #107130); UAS-InRRNAi

(VDRC stock #992); UAS-CNKRNAi (VDRC stock #11960); UAS-RASRNAi (VDRCstock #106642); UAS-RAFRNAi (VDRC stock #20909); UAS-MAPKRNAi (VDRCstock #43124). To generate a transgenic steppke RNAi line, a 443 bp stepcDNA fragment (bp 589–1031 of step-RA cDNA) was cloned in reverseorientation into pWiz via XbaI (using primers: 59-accaacacggacacatgcta-39 and59-gtcgcaaacaactcaaagca-39). Transgenic flies were generated via transposase-mediated P-element insertion. step knockdown of RNAi construct was tested usingreal-time RT-PCR and immunohistochemistry (see Fig. 1C–C0).

Rescue experiments were of the following genotype: hsFLP;stepK/stepK; act

.CD2 .Gal4, UAS-GFP/UAS-step.

To generate a steppke null allele, several gene targeting approaches wereperformed according to the modified ‘Rong and Golic’ protocol (Rong and Golic,2001; Huang et al., 2008). In contrast to parallel gene targeting approaches forother genes in the lab, several steppke specific targeting attempts failed, mostprobably due to the low recombination rate in the vicinity of the centromereregion. Moreover, Flp/Frt-based clonal analysis of steppke gene function using thehypomorphic alleles is prevented by the close localization of the steppke genelocus with respect to the FRT40a insertion.

Generation of a Steppke antibody

The coding sequence corresponding to amino acids 5 to 266 was cloned into thepTriEX-4 Neo vector via NcoI and Xho to generate an expression vector containingthe coding sequence for a 66His tagged Steppke fusion protein. Steppke5-266-HIS

was expressed in E. coli BL21 and purified via the 66HIS tag using NiNTAAgarose (Macherey-Nagel). Guinea pigs were immunized with the recombinantSteppke5-266-HIS protein. 2 mg of purified Steppke5-266-HIS was immobilized toCNBr activated Sepharose 4B in a column according to manufacturer’sinstructions (GE healthcare) to generate affinity purification columns. 2 mlserum of immunized animals was loaded to the column for 2 hours at roomtemperature. The column was further washed with PBT, and the purified antibodywas eluted via 0.1 M glycine (pH 2.3). Fractions were neutralized via Tris-HCl(pH 8.1) and amount and quantity of purified antibody was assessed viaCoomassie SDS page. Specificity of antibody was tested in step gain and loss offunction (Fig. 1B,C).

Immunohistochemistry

Larval imaginal discs were fixed in 4% paraformaldehyde/PBS for 45 minutes atroom temperature. Specimen were blocked for at least 1 hour in 1% RotiH-Block(Carl Roth)/PBT (containing 0.1% Triton X-100) and incubated with primaryantibodies in PBT over night at 4 C. Primary antibodies used were: anti-E-Cadherin (1:10, Santa-Cruz), anti-GFP (1:100, Sigma-Aldrich), anti-dpERK

Steppke controls EGFR/MAPK signaling 2477

Journ

alof

Cell

Scie

nce

(1:100, Sigma-Aldrich), anti-Argos (1:20, DSHB), anti-EGFR (1:100, gift fromB. Shilo), anti-CNK (1:50; a gift from M. Lefrancois), anti-Schlank (Bauer et al.,2009; Voelzmann and Bauer, 2011) and anti-Step (1:50). Alexa 488-, Alexa 633-and Cy3-conjugated (Invitrogen Molecular Probes) secondary antibodies wereused. Specimens were embedded using Fluoromount DAPI (SouthernBiotec).Images were obtained using a confocal microscope (Zeiss, LSM710) andprocessed in Adobe Photoshop. Co-localization was analyzed usingBioImageXD.

Quantitative real-time RT-PCR

For RNA isolation at least 40 imaginal wing discs or at least 10 Drosophila thirdinstar larvae were placed in lysis buffer and homogenized using an Ultra-TurraxT25basic. Total RNA was isolated by the NucleoSpin RNA II kit (Macherey-Nagel) and RNA concentration was analysed via a NanoDrop (ThermoScientific). For first strand cDNA reaction 500 ng of total RNA wastranscribed using the QuantiTect Reverse Transcription Kit (Qiagen) includingDNaseI treatment following the supplier’s protocol. Real-time PCR wasperformed with 1 ml cDNA per reaction using the IQ SYBR Green Supermix(Biorad) as detection dye. Experiments were performed with iQ5 Real-Time PCRDetection System from BIO-RAD. cDNA samples were run in triplicates, theaverage CT was used to analyse the expression levels via the 22DDCT method(Livak and Schmittgen, 2001). Experiments were repeated with independentlyisolated RNA samples from different egg collections. Actin 5C (Act5C, act) andRibosomal protein L32 (RpL32, rp49) were used as reference genes. PCR wasanalyzed using BIO-RAD iQ5 Optical System software (version 1.1.1442.OCR),following the instruction provided by the supplier and Microsoft Excel. Theoligonucleotides in supplementary material Table S1 were used for real-time PCRanalysis.

Wing analysis

Flies were anaesthetized; wings were dissected and placed on a slide with a drop ofwater. Wings were imaged under an Olympus SZ12 binocular via SIS Analysis 2.1software. Relative wing areas were analyzed via ImageJ (National Institute ofHealth).

Wing size was determined as ratio of posterior wing area (L3 was taken asborder) versus total wing areas (see supplementary material Fig. S1A). Relativewing cell size was assessed wing hair counts in a defined area of the wing (celldensity). Total cell number was derived from cell and wing size.

Yeast two-hybrid analysis

Full length and deletion variants of Steppke were cloned into pGBD-C3 (ClontechLaboratories, Inc.), CNK C-terminus (aa 818–1557) was cloned into pGAD-C1(Clontech Laboratories, Inc.). All constructs were confirmed by sequencing. Toanalyze the interaction of the CNK C-terminus with Steppke mutants in the two-hybrid assay, yeast strain AH109 (Clontech Laboratories) was transformed withcorresponding two-hybrid constructs. Expression of GAL4 DNA-binding domainand GAL4 activation domain fusion proteins was verified by immunoblottingusing monoclonal anti-GAL4 activation domain (sc-429, Santa CruzBiotechnology) and anti-GAL4 DNA-binding domain antibodies (sc-510, SantaCruz Biotechnology). Interaction of CNK C-terminus with Steppke fragments wasanalyzed by growth of corresponding transformants on synthetic dextrose minimalmedium lacking histidine, leucine and tryptophan for 7 days at 30 C.Cotransfection of pGAD-Chip and pGBD-N-Hsc70 was used as a positivecontrol, cotransfection of pGAD-Chip and pGBD-C-Hsc70 served as a negativecontrol (Arndt et al., 2005). Autoinduction of Step deletion constructs was testedby co-transformation of the empty pGAD-C1 vector. Autoinduction of the CNKconstruct was tested vice versa.

AcknowledgementsWe thank J. Hohfeld for sharing the pGAD-C1 and pGBD-C3vectors, the competent yeast strain AH109 and the pGBD-N-Hsc70,pGBD-C-Hsc70, pGAD-Chip constructs serving as positive andnegative controls in the yeast two-hybrid assay. We thank A.Volzmann for preparation of the Steppke5-266-HIS CNBr columnsand for discussions on the manuscript. We are grateful to B. Shilofor providing the anti-EGFR antibody, C. Klambt for the UAS-EGFR fly lines and M. Lefrancois for providing the anti-CNKantibody.

Author contributionsI.H. designed and performed most of the experiments andinterpreted data. B.F. designed experiments and interpreted data.A.P. contributed to the CNK wing experiments, T.W. contributed tothe analysis of EGFR target gene levels, A.P., T.W. and A.S.

performed the Y2H analysis. D.G. generated the Step antibody.M.H. planned and supervised the project. I.H., B.F. and M.H. wrotethe manuscript.

FundingThis work was funded by the Deutsche Forschungsgemeinschaft[grant numbers SFB704, SFB/TR83, SFB832 to M.H.] and a PhDfellowship to I.H. [grant number GRK804]. M.H. is a member of theBonn Excellence Cluster Immunosensation.

Supplementary material available online at

http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.120964/-/DC1

ReferencesAlvarez, R. H., Valero, V. and Hortobagyi, G. N. (2010). Emerging targeted therapies

for breast cancer. J. Clin. Oncol. 28, 3366-3379.

Antonarakis, E. S., Carducci, M. A. and Eisenberger, M. A. (2010). Novel targetedtherapeutics for metastatic castration-resistant prostate cancer. Cancer Lett. 291, 1-13.

Arndt, V., Daniel, C., Nastainczyk, W., Alberti, S. and Hohfeld, J. (2005). BAG-2acts as an inhibitor of the chaperone-associated ubiquitin ligase CHIP. Mol. Biol. Cell

16, 5891-5900.

Arkenau, H.-T. (2009). Gastric cancer in the era of molecularly targeted agents: currentdrug development strategies. J. Cancer Res. Clin. Oncol. 135, 855-866.

Bauer, R., Voelzmann, A., Breiden, B., Schepers, U., Farwanah, H., Hahn, I.,

Eckardt, F., Sandhoff, K. and Hoch, M. (2009). Schlank, a member of the ceramidesynthase family controls growth and body fat in Drosophila. EMBO J. 28, 3706-3716.

Becker, T., Loch, G., Beyer, M., Zinke, I., Aschenbrenner, A. C., Carrera, P.,

Inhester, T., Schultze, J. L. and Hoch, M. (2010). FOXO-dependent regulation ofinnate immune homeostasis. Nature 463, 369-373.

Bier, E. (2000). Drawing lines in the Drosophila wing: initiation of wing veindevelopment. Curr. Opin. Genet. Dev. 10, 393-398.

Bill, A., Schmitz, A., Albertoni, B., Song, J.-N., Heukamp, L. C., Walrafen, D.,

Thorwirth, F., Verveer, P. J., Zimmer, S., Meffert, L. et al. (2010). Cytohesins arecytoplasmic ErbB receptor activators. Cell 143, 201-211.

Blair, S. S. (2007). Wing vein patterning in Drosophila and the analysis of intercellularsignaling. Annu. Rev. Cell Dev. Biol. 23, 293-319.

Bohni, R., Riesgo-Escovar, J., Oldham, S., Brogiolo, W., Stocker, H., Andruss, B. F.,

Beckingham, K. and Hafen, E. (1999). Autonomous control of cell and organ sizeby CHICO, a Drosophila homolog of vertebrate IRS1-4. Cell 97, 865-875.

Britton, J. S., Lockwood, W. K., Li, L., Cohen, S. M. and Edgar, B. A. (2002).

Drosophila’s insulin/PI3-kinase pathway coordinates cellular metabolism withnutritional conditions. Dev. Cell 2, 239-249.

Brogiolo, W., Stocker, H., Ikeya, T., Rintelen, F., Fernandez, R. and Hafen,

E. (2001). An evolutionarily conserved function of the Drosophila insulin receptorand insulin-like peptides in growth control. Curr. Biol. 11, 213-221.

Brunet, A., Roux, D., Lenormand, P., Dowd, S., Keyse, S. and Pouyssegur, J. (1999).Nuclear translocation of p42/p44 mitogen-activated protein kinase is required forgrowth factor-induced gene expression and cell cycle entry. EMBO J. 18, 664-674.

Carnero, A. (2010). The PKB/AKT pathway in cancer. Curr. Pharm. Des. 16, 34-44.

Chen, R. H., Sarnecki, C. and Blenis, J. (1992). Nuclear localization and regulation oferk- and rsk-encoded protein kinases. Mol. Cell. Biol. 12, 915-927.

Claperon, A. and Therrien, M. (2007). KSR and CNK: two scaffolds regulating RAS-mediated RAF activation. Oncogene 26, 3143-3158.

Cobb, M. H. and Goldsmith, E. J. (2000). Dimerization in MAP-kinase signaling.Trends Biochem. Sci. 25, 7-9.

De Celis, J. F. (2003). Pattern formation in the Drosophila wing: The development ofthe veins. Bioessays 25, 443-451.

Dietzl, G., Chen, D., Schnorrer, F., Su, K.-C., Barinova, Y., Fellner, M., Gasser, B.,

Kinsey, K., Oppel, S., Scheiblauer, S. et al. (2007). A genome-wide transgenicRNAi library for conditional gene inactivation in Drosophila. Nature 448, 151-156.

Douziech, M., Roy, F., Laberge, G., Lefrancois, M., Armengod, A.-V. and Therrien,

M. (2003). Bimodal regulation of RAF by CNK in Drosophila. EMBO J. 22, 5068-5078.

Douziech, M., Sahmi, M., Laberge, G. and Therrien, M. (2006). A KSR/CNKcomplex mediated by HYP, a novel SAM domain-containing protein, regulates RAS-

dependent RAF activation in Drosophila. Genes Dev. 20, 807-819.

Fuss, B., Becker, T., Zinke, I. and Hoch, M. (2006). The cytohesin Steppke is essentialfor insulin signalling in Drosophila. Nature 444, 945-948.

Gabay, L., Scholz, H., Golembo, M., Klaes, A., Shilo, B. Z. and Klambt, C. (1996).EGF receptor signaling induces pointed P1 transcription and inactivates Yan proteinin the Drosophila embryonic ventral ectoderm. Development 122, 3355-3362.

Gaengel, K. and Mlodzik, M. (2003). Egfr signaling regulates ommatidial rotation andcell motility in the Drosophila eye via MAPK/Pnt signaling and the Ras effectorCanoe/AF6. Development 130, 5413-5423.

Hafner, M., Schmitz, A., Grune, I., Srivatsan, S. G., Paul, B., Kolanus, W., Quast,

T., Kremmer, E., Bauer, I. and Famulok, M. (2006). Inhibition of cytohesins bySecinH3 leads to hepatic insulin resistance. Nature 444, 941-944.

Huang, J., Zhou, W., Watson, A. M., Jan, Y.-N. and Hong, Y. (2008). Efficient ends-out gene targeting in Drosophila. Genetics 180, 703-707.

Journal of Cell Science 126 (11)2478

Journ

alof

Cell

Scie

nce

Junger, M. A., Rintelen, F., Stocker, H., Wasserman, J. D., Vegh, M., Radimerski,T., Greenberg, M. E. and Hafen, E. (2003). The Drosophila forkhead transcriptionfactor FOXO mediates the reduction in cell number associated with reduced insulinsignaling. J. Biol. 2, 20.

Klambt, C. (1993). The Drosophila gene pointed encodes two ETS-like proteins whichare involved in the development of the midline glial cells. Development 117, 163-176.

Kolanus, W. (2007). Guanine nucleotide exchange factors of the cytohesin family andtheir roles in signal transduction. Immunol. Rev. 218, 102-113.

Laberge, G., Douziech, M. and Therrien, M. (2005). Src42 binding activity regulatesDrosophila RAF by a novel CNK-dependent derepression mechanism. EMBO J. 24,487-498.

Lenormand, P., Brondello, J. M., Brunet, A. and Pouyssegur, J. (1998). Growthfactor-induced p42/p44 MAPK nuclear translocation and retention requires bothMAPK activation and neosynthesis of nuclear anchoring proteins. J. Cell Biol. 142,625-633.

Lim, J., Zhou, M., Veenstra, T. D. and Morrison, D. K. (2010). The CNK1 scaffoldbinds cytohesins and promotes insulin pathway signaling. Genes Dev. 24, 1496-1506.

Livak, K. J. and Schmittgen, T. D. (2001). Analysis of relative gene expression datausing real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25,402-408.

Mannell, H. K., Pircher, J., Chaudhry, D. I., Alig, S. K. C., Koch, E. G., Mettler, R.,

Pohl, U. and Krotz, F. (2012). ARNO regulates VEGF-dependent tissue responses bystabilizing endothelial VEGFR-2 surface expression. Cardiovasc. Res. 93, 111-119.

Martın-Blanco, E., Roch, F., Noll, E., Baonza, A., Duffy, J. B. and Perrimon,

N. (1999). A temporal switch in DER signaling controls the specification anddifferentiation of veins and interveins in the Drosophila wing. Development 126,5739-5747.

McNeill, H., Craig, G. M. and Bateman, J. M. (2008). Regulation of neurogenesis andepidermal growth factor receptor signaling by the insulin receptor/target of rapamycinpathway in Drosophila. Genetics 179, 843-853.

Rajakulendran, T., Sahmi, M., Kurinov, I., Tyers, M., Therrien, M. and Sicheri,F. (2008). CNK and HYP form a discrete dimer by their SAM domains to mediateRAF kinase signaling. Prxoc. Natl. Acad. Sci. USA 105, 2836-2841.

Rong, Y. S. and Golic, K. G. (2001). A targeted gene knockout in Drosophila. Genetics

157, 1307-1312.Roy, F., Laberge, G., Douziech, M., Ferland-McCollough, D. and Therrien,

M. (2002). KSR is a scaffold required for activation of the ERK/MAPK module.Genes Dev. 16, 427-438.

Schlessinger, J. (2004). Common and distinct elements in cellular signaling via EGFand FGF receptors. Science 306, 1506-1507.

Simpson, P. (2007). The stars and stripes of animal bodies: evolution of regulatory elementsmediating pigment and bristle patterns in Drosophila. Trends Genet. 23, 350-358.

Tabata, T., Schwartz, C., Gustavson, E., Ali, Z. and Kornberg, T. B. (1995). Creatinga Drosophila wing de novo, the role of engrailed, and the compartment borderhypothesis. Development 121, 3359-3369.

Teleman, A. A. (2010). Molecular mechanisms of metabolic regulation by insulin inDrosophila. Biochem. J. 425, 13-26.

Therrien, M., Wong, A. M. and Rubin, G. M. (1998). CNK, a RAF-bindingmultidomain protein required for RAS signaling. Cell 95, 343-353.

Voelzmann, A. and Bauer, R. (2011). Embryonic expression of Drosophila ceramidesynthase schlank in developing gut, CNS and PNS. Gene Expr. Patterns 11, 501-510.

Wasserman, J. D. and Freeman, M. (1997). Control of EGF receptor activation inDrosophila. Trends Cell Biol. 7, 431-436.

Wasserman, J. D. and Freeman, M. (1998). An autoregulatory cascade of EGFreceptor signaling patterns the Drosophila egg. Cell 95, 355-364.

Witte, H. T., Jeibmann, A., Klambt, C. and Paulus, W. (2009). Modeling gliomagrowth and invasion in Drosophila melanogaster. Neoplasia 11, 882-888.

Yoshida, H., Kwon, E., Hirose, F., Otsuki, K., Yamada, M. and Yamaguchi,

M. (2004). DREF is required for EGFR signalling during Drosophila wing veindevelopment. Genes Cells 9, 935-944.

Steppke controls EGFR/MAPK signaling 2479