Scholarly article on topic 'Sphingolipid depletion impairs endocytic traffic and inhibits Wingless signaling'

Sphingolipid depletion impairs endocytic traffic and inhibits Wingless signaling Academic research paper on "Biological sciences"

CC BY
0
0
Share paper
Academic journal
Mechanisms of Development
OECD Field of science
Keywords
{Wingless / Sphingolipid / Ceramide / "Endocytic trafficking"}

Abstract of research paper on Biological sciences, author of scientific article — Julia Pepperl, Gerlinde Reim, Ursula Lüthi, Andres Kaech, George Hausmann, et al.

Abstract Sphingolipids are an important part of the plasma membrane and implicated in a multitude of cellular processes. However, little is known about the role of sphingolipids in an epithelial context and their potential influence on the activity of signaling pathways. To shed light on these aspects we analyzed the consequences of changing ceramide levels in vivo in the Drosophila wing disc: an epithelial tissue in which the most fundamental signaling pathways, including the Wnt/Wg signaling pathway, are well characterized. We found that downregulation of Drosophila’s only ceramide synthase gene schlank led to defects in the endosomal trafficking of proteins. One of the affected proteins is the Wnt ligand Wingless (Wg) that accumulated. Unexpectedly, although Wg protein levels were raised, signaling activity of the Wg pathway was impaired. Recent work has spotlighted the central role of the endocytic trafficking in the transduction of the Wnt signal. Our results underscore this and support the view that sphingolipid levels are crucial in orchestrating epithelial endocytic trafficking in vivo. They further demonstrate that ceramide/sphingolipid levels can affect Wnt signaling.

Academic research paper on topic "Sphingolipid depletion impairs endocytic traffic and inhibits Wingless signaling"

Available at www.sciencedirect.com

SciVerse ScienceDirect

journal homepage: www.elsevier.com/locate/modo

chanisms

Develi

Sphingolipid depletion impairs endocytic traffic and inhibits Wingless signaling 5

Julia Pepperl a, Gerlinde Reim a, Ursula Liithi b, Andres Kaech b, George Hausmann a, Konrad Basler a'*

a Institute of Molecular Life Sciences, University of Zurich, Zurich, Switzerland b Center for Microscopy and Image Analysis, University of Zurich, Zurich, Switzerland

ARTICLE INFO

ABSTRACT

Article history: Received 12 March 2013 Received in revised form 23 April 2013 Accepted 29 April 2013 Available online 9 May 2013

Keywords: Wingless Sphingolipid Ceramide

Endocytic trafficking

Sphingolipids are an important part of the plasma membrane and implicated in a multitude of cellular processes. However, little is known about the role of sphingolipids in an epithelial context and their potential influence on the activity of signaling pathways. To shed light on these aspects we analyzed the consequences of changing ceramide levels in vivo in the Drosophila wing disc: an epithelial tissue in which the most fundamental signaling pathways, including the Wnt/Wg signaling pathway, are well characterized. We found that downregulation of Drosophila's only ceramide synthase gene schlank led to defects in the endosomal trafficking of proteins. One of the affected proteins is the Wnt ligand Wingless (Wg) that accumulated. Unexpectedly, although Wg protein levels were raised, signaling activity of the Wg pathway was impaired. Recent work has spotlighted the central role of the endocytic trafficking in the transduction of the Wnt signal. Our results underscore this and support the view that sphingolipid levels are crucial in orchestrating epithelial endocytic trafficking in vivo. They further demonstrate that ceramide/ sphingolipid levels can affect Wnt signaling.

© 2013 The Authors. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Sphingolipids influence a variety of developmental processes including cell growth, differentiation and apoptosis. They are synthesized in the ER and Golgi apparatus and are subsequently transported to the plasma membrane, where they reside in the external membrane leaflet (Futerman and Riezman, 2005). Sphingolipids were proposed to form the controversial detergent resistant microdomains, the so-called "lipid rafts'', together with cholesterol. These microdomains within the plasma membrane form, as sphingolipids tend to cluster due to lateral interactions; additionally cholesterol

intercalates between their mostly saturated fatty acids, giving rise to a more rigid or "liquid-ordered" patch of membrane. Due to the differential membrane composition some lipid-modified and transmembrane proteins might preferentially associate with these lipid microdomains, whereas others would avoid them (Lingwood and Simons, 2010; Simons and Ikonen, 1997). This led to the notion that sphingolipids are involved in protein sorting and trafficking (Le Roy and Wrana, 2005). It also suggested a potential role in regulating signaling activity by providing platforms for signaling complexes (Simons and Toomre, 2000). However, the effect of changing sphingolipid levels in an epithelial context, for example the

5 This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. * Corresponding author. E-mail address: basler@imls.uzh.ch (K. Basler). 0925-4773/$ - see front matter © 2013 The Authors. Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.mod.2013.04.001

consequences for signaling pathway activities, is not well understood. Ceramide, the precursor of all complex sphin-golipids, is known to be an important second messenger, as well as a modulator of membrane properties. Ceramide production is induced during various stress situations and is thought to be involved not only in apoptotic signaling (Kolesnick and Kronke, 1998), but also in cell growth and differentiation. Due to its physicochemical properties cera-mide tends to lateral clustering which, combined with its inverted conical shape, results in negative membrane curvature and inwards budding of membrane (Holopainen et al., 2000). It is still debated if ceramide functions directly as a second messenger or indirectly via its ability to induce changes in the membrane structure (Kolesnick et al., 2000; Van Blitterswijk et al., 2003). This influence on membrane structure further suggests a role of ceramide in endocytosis and intracellular trafficking: when fibroblasts and macrophages are treated with an exogenous sphingomyelinase in the absence of ATP the formation of endocytic vesicles was observed (Zha et al., 1998).

There is growing evidence that endocytosis and endocytic trafficking of the ligand-receptor-complexes are essential for Wnt/Wg signal transduction both in mammalian cells and in Drosophiia (Blitzer and Nusse, 2006; Seto and Bellen, 2006). This highly conserved signaling pathway plays a crucial role in a variety of developmental processes and deregulated Wnt signaling is implicated in various diseases, including colon cancer and degenerative diseases (Clevers and Nusse, 2012). In the canonical branch of Wnt/Wg signaling Wnt forms a ternary complex with its receptors Frizzled (Fz) and LRP5/6 (Arrow in Drosophiia), which are activated by a series of phosphorylation events. This leads to relocalization of Dishevelled (Dvl) to the plasma membrane and the formation of receptor complexes on oligomerized Dvl clusters, giving rise to signaling platforms called "signalosomes" (Bilic et al., 2007). The binding of the scaffold protein Axin leads to the recruitment of APC and GSK3b, which are also components of the b-catenin destruction complex. As a consequence b-catenin can no longer be degraded, accumulates and translocates to the nucleus, where it initiates the transcriptional activation of target genes. Several observations highlight the importance of endocytic trafficking in Wnt signaling: first, the Fz receptor stimulates its own endocytosis via Gao and the early-endosomal GTPase Rab5, and localization in the respective endosomal compartment is reported to influence the balance between canonical and non-canonical signaling branch (Purvanov et al., 2010). Second, Niehrs and colleagues showed that acidification of an intracellular compartment is required for LRP6 receptor phosphorylation and thus Wnt signaling induction (Buechling et al., 2010; Cruciat et al., 2010). Third, it was shown in an elegant study that GSK3b is sequestered into multivesicular bodies (MVBs) upon Wnt stimulation, so that the enzyme is secluded from its cytosolic substrate b-catenin and by this the Wnt signal can be transduced (Taelman et al., 2010).

Given the dearth of knowledge about the role of sphingoli-pids in an epithelial context and their potential influence on the activity of signaling pathways we set out to analyze the consequences of changed ceramide levels, using the Drosoph-

iia wing disc as a suitable model for an epithelium. Since most fundamental signaling pathways, including the Wnt/Wg signaling pathway, are well characterized in the wing imaginal disc we hoped to gain insights into the effects of ceramide depletion on these pathways. To this end, we used RNAi-med-iated gene knockdown against components of the de novo ceramide synthesis pathway (Acharya and Acharya, 2005). Our study focuses on the effects of depleting ceramide levels by downregulation of schiank, the only ceramide synthase gene in Drosophiia. We found that downregulation of schiank does not obviously influence endocytosis and formation of the early-endosomal compartment. It rather leads to disruption of recycling and degradative endocytic trafficking routes and consequently the accumulation of several secreted and transmembrane proteins in the cells, among them Wg. In spite of the higher Wg levels, the activity of the Wg signaling pathway is reduced. This reduction of Wg signaling activity is possibly connected to an early-to-late endosomal trafficking defect and highlights the importance of endocytic trafficking for Wg signal transduction. Importantly, the results reveal a hitherto unknown function for sphingolipids in Wnt signaling.

2. Results

2.1. Depletion of schlank leads to growth defects, accumulation of Wg protein and reduced Wg signaling activity

To determine the effects of sphingolipid depletion in an epithelial context, we examined the effect of reducing schlank function (CG3576, Bauer et al., 2009). This choice was based on the facts that its product is Drosophila's only ceramide synthase and that its downregulation results in reduced ceramide levels, and on the availability of suitable alleles. Schlank was previously identified in a genome-wide RNAi screen conducted in our lab to uncover growth regulators (G. Reim and K. Basler, unpublished). Consistent with this, schlank depletion by two independent RNAi lines (Fig. 1A) in the posterior compartment resulted in reduced cell size, as judged by the closer arrangement of wing trichomes (Fig. 1B''), and increased levels of apoptosis (Fig. S1A). The wing margins were frequently found to be lost (Fig. 1B and Fig. S1C), a phenotype characteristic of reduced Wg signaling. Furthermore, the proximo-distal orientation of the trichomes was disturbed (Fig. 1B''), reminiscent of defects in the non-canonical Wnt or planar cell polarity (PCP) pathway. To begin to explain these observations, we monitored Wg distribution and analyzed the expression of Wg target genes. Wg protein was massively accumulating in dot-like structures upon schlank depletion (Fig. 1C). Whereas expression of the low-threshold target gene Distalless (Dll) was not changed (data not shown), the expression of the high-threshold target gene Senseless (Sens) was reduced (Fig. 1D, Fig. S1B and D).

These data indicate that besides the growth phenotypes, reduction of ceramide levels also leads to accumulation of Wg protein in vesicle-like structures. Unexpectedly although there is more Wg ligand, high-level Wg signaling activity seems to be impaired upon schlank depletion.

A P365

\7 RNAH

schlank

1000 bp RNAi2

-□□am

RNAH hh::schlank

ilaiiigfsfc

' , ,,

hh::schlank

RNAi2 ■ ' RNAH

^ : hh::sch/ank

<f •A y ■ *

Ir' i, l. *. :

m iT* v^ >*

GFP hh. RNAH ;:schlank

_* - - jmr .

-if*. ■

GFP hh::schlank

• < . —-—

• ■/ /

Sens " ■ . .

Fig. 1 - Phenotypes upon schlank depletion. Overview of the schlank gene region including the alleles used and generated in this study (A). Adult wings showed notches and disturbed wing hair orientation upon expression of two independent RNAi lines under the control of hh-Gal4 in the posterior compartment (B). In third instar larval wing imaginal discs the RNAi expression with hh-Gal4 in the posterior compartment, marked by the expression of UAS-GFP, showed Wg protein accumulation as judged by Wg antibody staining (C) and reduced Wg signaling activity, detected by Sens antibody staining (D). Scale bars represent 50 im.

2.2. Wg accumulation is due to impaired ceramide synthase activity and independent of apoptosis

To confirm that the observed phenotypes are a consequence of downregulation of schlank and to exclude potential RNAi mediated off-target effects, we tested the hypomorphic P-element-based allele schlankP365, in which the P-element is inserted downstream of the first schlank exon. We observed the same Wg accumulation phenotype as in schlank RNAi

expressing wing discs (data not shown). Additionally, we generated a presumptive null allele, schlank331, via imprecise excision of this P-element, deleting the downstream half of the first exon including a predicted transmembrane domain (Fig. 1A). Again, we observed the cell-autonomous accumulation of Wg protein in schlank331 clones (Fig. 2A). Due to the impaired growth upon schlank depletion, we mostly worked in a Minute background to equip the schlank mutant cells with a growth advantage. These experiments show that the

observed defects are caused by the depletion of schlank and are not an RNAi artifact.

We next performed rescue experiments using an actin-Gal4 driven UAS-schlank transgene (either native or HA-tagged) and a genomic construct comprising the entire schlank locus. All of these constructs were able to fullyrescue either schlankP365 or schlank331 mutant flies. In a wild-type background they did not cause any phenotype. Importantly, we also tested the allele H215D, in which the enzymatic activity of Schlank is abrogated (Bauer et al., 2009), and found it unable to rescue our schlank alleles (data not shown). RNAi induced Wg accumulation was diminished by expression of Schlank but not the Schlank mutant H215D (Fig. S2A). We further induced schlankP365 mutant clones and expressed the HA-tagged Schlank protein in the posterior compartment: again the schlankP365 induced Wg accumulation was reverted (Fig. 2B). The rescue capability of Schlank and the inability of the H215D mutant strongly suggest that the Wg accumulation is due to reduced ceramide synthase activity.

To test if the Wg accumulation is merely a consequence of elevated levels of apoptosis, we provoked apoptosis by expressing the pro-apoptotic protein Hid in the posterior compartment. In this case, the apoptosis induction (Fig. S2B) did not result in Wg accumulation (Fig. 2C). Next, we prevented apoptosis in schlank depleted cells by expressing Diapl protein in the posterior compartment (Fig. S2C) and found Wg protein still accumulated (Fig. 2D). Hence this phenotype is not a consequence of the elevated levels of apoptosis upon schlank reduction.

Taken together these experiments indicate that the Wg accumulation phenotype is caused by reduced ceramide syn-thase activity and is independent of apoptosis.

2.3. Several secreted and transmembrane proteins accumulate, but their respective signaling pathway activities are not changed

To test if the reduction of ceramide levels affects wing imaginal disc cell morphology, we assessed the overall integrity of the wing disc epithelium by electron microscope analysis of wing discs expressing schlankRNAl1 in the posterior compartment. For visualization of these cells we additionally expressed a membrane-bound Horseradish Peroxidase that could be detected via DAB staining. We were able to identify schlank mutant tissue using this method and found that the apical-basal cell morphology and the tissue integrity of schlank depleted cells was not disrupted (Fig. S2D).

Next we asked if Wg was the only protein accumulating as a consequence of schlank loss/reduction of function. We tested several other secreted and transmembrane proteins for accumulation upon schlank depletion. Antibody stainings against the Wg receptor Fz2 and its co-receptor Arrow, as well as Hedgehog (Hh), its receptor Patched, and the surface receptor Notch all revealed the same dot-like accumulation phenotype (Fig. 3, data not shown). The cell adhesion protein E-cadherin also accumulated but to a much lower extent (Fig. S3A). As there are no suitable antibodies for Dpp available, we expressed a GFP-tagged Dpp protein in the dpp expression domain. Interestingly, this did not show accumulation (Fig. S3B). The different behavior could be due to different endocytic trafficking routes of Dpp or reduced protein turnover in the case of E-cadherin.

We further examined if the signaling activity of the Hh, Notch and Dpp pathways was changed. To address this we made use of various readouts for the respective pathways including antibody stainings, stainings for lacZ-reporters of target genes and rtPCR analysis (Fig. S3C and D, data not shown). In contrast to the effect on Wg signaling, schlank depletion did not detectably change the activity of the Hh, Notch and Dpp signaling pathways.

In summary, we found that schlank depletion does not grossly affect the overall integrity of the wing disc epithelium. However it does cause the accumulation of a subset of secreted and transmembrane proteins. The extent of the accumulation varied, but with the exception of Wg there appeared to be no detectable changes in the activity of the signaling pathways. Therefore we decided to concentrate our attention on the Wg signaling pathway.

2.4. Wg accumulation is not due to more production and secretion

Two possible scenarios could account for the observed Wg accumulation: first, the accumulating Wg could be due to more production and secretion of the protein. Second, there could be less degradation of Wg protein. The cell autonomy of the observed phenotype already hinted to less degradation as a secretion defect would be expected to be non-autonomous. To test this in more detail, we monitored wg transcription using a lacZ-reporter and performed extracellular Wg stainings, which constitute a very sensitive readout for Wg secretion. Neither the transcription of wg (Fig. 4A) nor the extracellular level of the Wg protein was changed upon schlank depletion (Fig. 4B and C), demonstrating that the observed accumulation is not caused by more production and secretion of the Wg protein. Furthermore we did not detect changes in the uptake of fluo-rescently labeled Dextran in schlank depleted cells, suggesting that the initial step of endocytosis is not disturbed (data not shown). Thus the Wg accumulation is probably due to impaired degradation after the protein was taken up.

Consistent with the hypothesis that upon schlank depletion there is a defect in endosomal trafficking we observed that the typical punctate localization of Wntless in the Golgi apparatus in Wg-producing cells was disrupted (Fig. 4D). Wg secretion is dependent on the Wntless/Evi (Wls) protein, which promotes Wg release and is afterwards recycled to the Golgi apparatus in a retromer-dependent manner (Port et al., 2008). Wls was apparently not as efficiently recycled back to the Golgi apparatus upon schlank depletion, whereas Wg secretion seemed unchanged. Therefore, we speculate that the amount of Wls in the Golgi apparatus is still sufficient to ensure proper Wg secretion, but cannot rule out secondary effects as, e.g., a change in the potency of the secreted ligand to induce Wg signaling (Franch-Marro et al., 2008).

2.5. Schlank depletion leads to enlarged Rab4/7/11 positive endosomal structures

To see where the endosomal trafficking is defective, we made use of various Rab GTPases as endosomal markers. Rab5 GTPase marks early endosomal structures, whereas Rab4 and Rab11 GTPases label the fast and slow recycling

i '0 "

ß-Gal schlan^^^ en::diap Gal80

Fig. 2 - Wg accumulation is indeed caused by schlank reduction and independent of apoptosis. (A) schlank331 mutant clones, marked by the absence of GFP, were induced by mitotic recombination in a Minute background and also displayed accumulating Wg protein. (B) Providing back HA-tagged Schlank protein via hh-Gal4 in analogous generated schlankp365 clones rescued the Wg accumulation specifically in the posterior compartment (compare arrows). (C) En-Gal4 driven expression of UAS-Hid in the GFP-positive P-compartment was induced 24 h after egg-laying by temperature-shift to provoke apoptosis. Staining for Wg did not show protein accumulation. (D) Rescue of apoptosis by en-Gal4 mediated Diapl expression 72 h after egg-laying did not diminish Wg protein accumulation in schlankp365 clones, which were marked by absence of ß-Gal staining. Scale bars represent 50 im.

Fig. 3 - Several proteins accumulate upon reduced schlank expression. GFP-negative schlankp365 clones were induced by mitotic recombination in a Minute background and stained with antibodies against Fz2 (A), Arrow (B), Hh (C) and Ptc (D). In all of these cases the proteins showed accumulation in the clones. Scale bars represent 50 im.

endosomes, respectively. Rab7 is used to label the late endosomal trafficking route leading to lysosomal degradation. Constructs consisting of the tubulin promoter driven Rab GTPase fused with an YFP-tag were employed and schlank clones were induced in this background. While Rab5 positive

endosomal compartments were not detectably altered by schlank depletion (Fig. 5A), Rab4/7/11 positive endosomes were massively enlarged (Fig. 5B-D). To check if other intracellular compartments were affected, we analyzed the status of Golgi apparatus and Endoplasmic Reticulum (ER) by testing the

Golgi component b-Cop and ER-resident KDEL-carrying proteins. We could not detect any differences in the antibody sta-inings against these proteins (Fig. S4A and B).

These experiments show that the processes of endocytosis or early endosome formation are not so sensitive to alteration in sphingolipid levels, in contrast to the more downstream intracellular trafficking steps.

2.6. All endosomal compartments exhibit increased Wg concentration

Given that we see accumulation of ligand (Wg) and receptor (Arr, Fz2) we were surprised to find that Wg signaling was reduced. Recent work on the Wg signaling transduction mechanism offers two possible explanations for this observation in the context of endocytic trafficking: first, it was shown that Wg has to enter the Rab5 positive early endosome for full strength signaling (Purvanov et al., 2010). It is possible that, due to defects in endocytic sorting and traveling, there is in absolute terms less Wg in Rab5 positive compartments, which is masked by Wg accumulation in other endosomal compartments. Alternatively, the underlying defect could be occurring in the more downstream steps of the degradative route such as impaired lysosomal acidification, which is essential for Wg signal transduction (Buechling et al., 2010; Cruciat et al., 2010), or due to impaired formation of multive-sicular bodies, which are reportedly necessary for GSK3b sequestration in vertebrate systems (Taelman et al., 2010).

To test if less Wg in Rab5 early endosomes could account for the Wg signaling defect, we quantified the Wg protein in Rab5 early endosomes, as well as in Rab4 recycling and in Rab7 late endosomes. As Wg endocytosis is believed to be receptor-mediated (Dubois et al., 2001) and by staining for Frizzled receptor we would not be able to distinguish between ligand-bound endocytosed receptor and "empty" recycling receptor, we used Wg staining to detect endocytosed ligand-receptor complex. As markers of the endosomal compartments, we again used the YFP-Rab constructs, now in combination with schlank RNAi driven in the posterior compartment. In this setup, we first defined the endosome volumes in control and schlank depleted compartments in the same disc to test how much the different Rab-positive compartments are enlarged upon schlank depletion. Next, we quantified the amount of Wg in the mutant and the neighboring control compartment. This yielded quantitative information about how much Wg is accumulating overall in the schlank depleted situation. Third, we measured the amount of Wg that could be found inside a particular type of endosome. To do so, we analyzed the ratio of Wg signal in mutant compared to control endosomes and normalized for the endosomal volume.

Confirming our microscopic analysis of the respective Rabpositive endosomal compartments, we did not see an enlargement of the Rab5-marked endosomal compartments. In contrast, we observed a twofold enlargement of Rab4-posi-tive endosomal structures and an even more pronounced enlargement of Rab7 late endosomal structures (Fig. 6A). Overall we found a twofold accumulation of Wg protein upon schlank depletion. Rab4- and Rab5-positive compartments showed a twofold increase in the concentration of Wg. In

Rab7-positive endosomes, more than fourfold more Wg was measured upon schlank reduction (Fig. 6B).

These results indicate that Wg protein levels in Rab5-posi-tive endosomes are increased, rather than decreased, upon schlank depletion (also when measured in absolute amounts and not corrected for the slightly decreased Rab5 endosomal volume), ruling out the notion that less Wg protein in Rab5 early endosomes is the cause for the reduced signaling activity.

Based on current models of Wnt/Wg signaling, a defect more downstream in the early-to-late endosomal pathway could result in reduced Wg signal transduction. The ligand-receptor complexes are thought to be sorted in a Hrs-depen-dent manner from Rab5-positive early endosome into multivesicular endosomes (Taelman et al., 2010). Therefore we analyzed the localization of Hrs and Vps16, both markers of multivesicular bodies, in our schlank reduced background. The localization of neither Hrs nor Vps16 was changed upon schlank depletion (Figs. 6C and S4C). In contrast, Lampl-GFP and LysoTracker, markers of the lysosomal compartment were enlarged in the schlank depleted situation (Figs. 6D and S4D).

Taken together, the observed protein accumulation and the defects in the formation of Rab7 endosomes and lyso-somes suggest that ceramide depletion impedes Wg signaling by disrupting the early-to-late endosomal route.

3. Discussion

3.1. Sphingolipids and endosomal trafficking

In this study, we reduced sphingolipid levels in the Dro-sophila wing imaginai discs by genetic means to study the effects of changed ceramide availability, with a focus on the activity of signaling pathways in an epithelial context.

Not unexpected given the coordination of energy metabolism and growth with lipid metabolism, we observed reduced cell size and cell number due to apoptosis upon depletion of the ceramide synthase gene schlank. When we examined the Drosophila wing imaginal disc we found that overall tissue integrity was not affected. However, we did observe an accumulation of various secreted and transmembrane proteins. Analysis of the endocytic compartments revealed that the slow and fast recycling endocytic routes, as well as the late endosomal route, were disrupted. In contrast, exocytosis (secretion of Wg and Dpp), the initial steps of endocytosis and the formation of Rab5-positive early endosomal compartments were not affected.

Ceramide is able to induce membrane curvature in artificial membranes (Holopainen et al., 2000). Additionally, changes in membrane morphology and fusion were reported in lipids extracted from mice defective in long chain ceramide synthesis (Silva et al., 2012). Thus, via physical changes in membrane budding/fusion sphingolipid composition could influence endocytic trafficking.

Consistent with a problem in the early-to-late endosomal route leading to lysosomal protein degradation we observed the accumulation of various proteins upon ceramide syn-thase depletion. These included Hedgehog (Hh), its receptor Patched and the surface receptor Notch (Fig. 3). We saw no effect on the Hh or Notch signaling pathways, whereas earlier

RNAH m

GFP 1ivgZ h'h::schlank lli-Gal

a..... B

I gy - 'flr^fijaaaWH *■',* ¿¡KmMfB

Jflik t A L M

GFP . • RNAH hh::schlSnk

B" js^l

Merge i - " ■

Fig. 4 - Wg accumulation is due to less degradation instead of more production. Staining against p-Gal of a wg-IacZ third instar wing imaginal disc expressing schlank RNAi with hh-Gal4 in the posterior compartment indicated unchanged wg transcription (A). (B) Detection of secreted Wg via antibody staining specifically for extracellular protein did not show differences upon posterior schlank RNAi expression. (C) Consistently, extracellular Wg staining was also not changed in schlankp365 clones induced in Minute background. (D) In the same genetic setup, staining with an Wntless antibody revealed diminished Golgi localization of Wntless/Evi (Wls) upon schlank depletion. Scale bars represent 50 im.

studies in Drosophila implicated glycosphingolipids in various steps in development and different signal pathways (Kraut, 2011; Singh et al., 2011). For example, in an impressive study

Hamel and colleagues show that glycosphingolipids modulate the signaling activity of Notch ligands (Hamel et al., 2010). The fact that we do not find reduced signaling activity apart from

ß-Gal schlank365 tubYFPRab11

Fig. 5 - Schlank depletion results in enlarged Rab4/7/11-positive endosomal structures Tubulin promoter driven, YFP-tagged Rab5- (A), Rab4- (B), Rab7- (C) and Rab11-constructs (D) were combined with schlankP365 mitotic clones to monitor the individual endosomal compartments. SchlankP365 clones were marked by the absence of p-Gal staining. Whereas Rab5-positive endosomes seemed to be unchanged, Rab4/7/11-positive endosomal structures were enlarged in schlankP365 clones. Scale bars represent 50 im.

Wg signaling suggests that sufficient glycosphingolipids are still produced, either by the remaining de novo synthesized ceramide or via salvage pathways. Taking this into account, our results show that the Wnt/Wg signaling pathway is particularly sensitive to the reduction in ceramide levels.

3.2. Sphingolipids and Wingless signaling

One of the many proteins that accumulated was the mor-phogen Wg. Although there was more Wg ligand in the system, the pathway activity was reduced upon schlank

• 'V • {

* vj* ÍL

Merge _

Fig. 6 - Wg accumulates preferentially in early-to-late endosomal route upon schlank reduction. (A) Wing imaginal discs carrying the above-mentioned YFP-Rab constructs and expressing hh-Gal4 driven schlank RNAi in the posterior compartment were employed to quantify the volumes of Rab 4/5/7-positive endosomal compartments. Using the automated surface creation tool on a defined cubicle volume in control and mutant compartment the size of the respective Rab endosomes was measured as ratio between mutant to control volumes in every disc. N = 7 for each genotype, P-values paired t-Test: Rab4: 0.0003, Rab5: 0.01, Rab7: 0.0002. (B) To quantify the overall amount of Wg accumulation independent of its endosomal localization, we assessed the fluorescence intensity ratio of a Wg antibody staining in the cubicles in mutant versus control compartment (N = 21). Following, we defined the concentration of Wg in the distinct endosomal compartments by applying the same method to the previously defined endosomal volumes and normalizing for the endosomal volume. N = 7 for each genotype, P-values paired t-Test: Rab4: 0.01, Rab5: 0.00002, Rab7: 0.0001. (C) To test more components of the early-to-late endosomal route we performed antibody stainings for Hrs in discs harboring schlankp365 clones, which showed no difference. (D) In the same genetic setup, we analyzed lysosomes using the LysoTracker® reagent and found the signal to be enlarged in schlankp365 clones. Scale bars represent 50 im.

depletion. We showed that this effect is dependent on cera-mide synthase activity and is not a secondary effect of apop-tosis. Furthermore, we demonstrated that the accumulation of Wg is not due to increased expression or secretion and happens in the late endosomal route and to a lesser extent in the recycling endosomes.

The Wg signaling defect that we see could result from impaired protein sorting. Sphingolipids are, beside their implied

role in membrane morphology, thought to be involved in protein sorting (Le Roy and Wrana, 2005; Lippincott-Schwartz and Phair, 2010; Simons and Ikonen, 1997). The partitioning or differential sorting of proteins into lipid microdomains is implicated in promoting the assembly of signaling complexes. In this regard, it is interesting that the Arrow homolog LRP6 is associated with membrane domains composed of cholesterol and sphingolipids (Yamamoto et al., 2006). A simple scenario

is that this association is important for signaling by promoting receptor clustering. Interfering with or abolishing this clustering could be envisaged to result in reduced signaling efficiency. However, given the growing body of evidence that connects endosomal trafficking and Wnt/Wg signaling, a parsimonious explanation is that a disruption of endosomal routes is responsible for the observed effects on the Wg pathway.

The first link between Wg signaling activity and internalization was made in 2006 (Blitzer and Nusse, 2006; Seto and Bellen, 2006). Later in an elegant study by the Niehrs lab it was shown that in the mammalian system Wnts induce clustering of receptors and Dvl in endocytosed complexes called signalo-somes, which facilitates pathway induction (Bilic et al., 2007). More recently a series of studies has highlighted various aspects of the relationship between Wnt signaling and endosomal trafficking: it could be shown that localization of Wg in Rab5-positive endosomal compartments is required for full-strength signal induction (Purvanov et al., 2010). Furthermore, acidification occurring along the early-to-late endosomal route was demonstrated to be important for Wnt signaling (Cruciat et al., 2010). Finally, it was reported that the GSK3b kinase has to be sequestered into multivesicular endosomes for proper Wnt signaling induction (Taelman et al., 2010).

Given that reduced schlank expression leads to defects in endosomal trafficking and Wnt/Wg signaling is critically linked to endosomal transport it is tempting to suggest that reduced Wg signaling we observed upon schlank depletion is caused by the impairment of endocytic trafficking. To further clarify the connection of Wnt signaling and ceramide levels, a cell-biological system has to be developed that provides the necessary resolution to dissect endocytic pathways during Wnt signaling upon ceramide depletion.

In summary, we present evidence that reducing sphingo-lipid levels in an epithelial context leads to the impairment of distinct endocytic routes. Furthermore, we show that the Wg signaling pathway is particularly sensitive to changes in cellular ceramide levels, and we believe this to be a consequence of defects at or below the level of early-to-late trafficking. This is the first demonstration of a connection between ceramide/ sphingolipid levels and Wg signaling and highlights the importance of these lipids in regulation of signaling processes.

4. Experimental procedures

4.1. Drosophila strains

4.1.1. schlank alleles

schlank RNAi1: VDRC 33896 (III), schlank RNAi2: VDRC 109418 (II), sch!ankP365: Bloomington 11985, schíank331: imprecise excision of schlankP365

4.1.2. Transgenes sch!ankHA/sch!ankH215DHA/genomic rescue: all constructs

were inserted in ZH-86Fb landing site via the /C31 integrase system.

4.1.3. Other stocks

ywf; Sp/CyO; hhGa!4/Tm6b, y wf;; act>CD2>Ga!4 UAS-GFP, y w arm-!acZ FRT19 hsp-flp (/FM7a), ywf; Sp/CyO; UAS-hid/Tm6b, ywf; Sp/CyO; UAS-diap1.

In general, we analyzed male larvae for dissection in RNAi experiments and female larvae for schlankP365/331 clonal analyses, as schlank is located on the X chromosome.

Schlank RNAi clones were generated by crossing the respective schlank RNAi line to ywf;; act>CD2>Gal4 UAS-GFP. Larvae were heat shocked 48 h after egg laying (AE) for 30 min at 37 °C. Schlank331 (and analogous schlankP365) Minute Rps5a clones were induced by heat shocking yw schlank331 FRT19/yw ubi-GFPnls Min FRT19 female larvae 48 h AE for 30 min at 37 °C. For the schlankP365 clonal rescue experiment flies with the genotype yw schlankP365 FRT19/yw ubi-GFPnls Min FRT19;; schlankHA hhGal4/+ were used to induce clones in the above mentioned Minute background. Rescue of schlank RNAi was done in ywf; Sp or CyO/+; schlankHA hhGal4/schlank RNAi1 flies. For combining schlankP365 clones with various Rab endosomal markers, the following flies were heat shocked as described above: yw schlankP365 FRT19/yw arm-lacZ FRT19 hsp-flp; tuba-YFP-Rab4/5/7/11/+. To test Dpp accumulation, yw schlankP365 FRT19/yw arm-lacZ FRT19 hsp-flp;; dppGal4 Gal80ts GFPdpp/+ flies were heat shocked 48 h AE (30 min, 37 °C) and GFP-Dpp expression was induced by temperature shift to 29 °C 96 h AE. Apoptosis induction was done in ywf; Sp/enGal4 Gal80ts UAS-CD8-GFP; UAS-hid/MKRS flies that were shifted to 29 °C 24 h AE. Rescue of apoptosis in schlankP365 clones was performed in the following genetic setup: yw sch-lankP365 FRT19/yw arm-lacZ FRT19 hsp-flp; enGal4 Gal80ts UAS-CD8-GFP/+; UAS-diap1/+. SchlankP365 clones were induced 48 h AE and diapl expression by shift to 29 °C 72 h AE.

4.2. Cloning and transgene production

The Lag1 open-reading frame was based on amplification of the cDNA clone LD18904 (BDGP) and cloned into a pUASattB vector for C-terminal 3xHA tagging. For the ceramide syn-thase dead version H215D (Bauer et al., 2009) we performed site-directed mutagenesis (QuikChange™ site-directed muta-genesis kit, Agilent Technologies) on a EcoRI/BamHI-sub-cloned fragment of the schlankHA construct using the primers:fwd CTGGCAGATGTTCATCgATCACATGGTCACCCTG CTCCTAATGrev CATTAGGAGCAGGGTGACCATGTGATCGATGA ACATCTGCCAG.

The mutated fragment was then swapped back in the pattB-UASschlankHA vector again via EcoRI/BamHI.

The genomic rescue fragment was based on BAC CH322-128M11 from the P[acman] library BACPAC Resources Center.

4.3. Generation of presumptive null allele sch!ank331

For the schlank331 allele, the Bloomington schlankP365 fly stock was used to generate an imprecise excision. This deleted the last 63 base pairs of the first schlank exon (including a predicted transmembrane domain) and fused the intronic region downstream of the P-element P365 insertion site to the remaining first exon.

4.4. Electron microscopy

For analysis, male larvae with the genotype ywf; UAS-CD2-HRP/Sp or CyO; schlank RNAi1/hhGal4 were dissected and immediately fixed with 2.5% glutaraldehyde in PBS for

15 min. After washing in PBS they were stained for 2 min using the DAB Peroxidase Substrate Kit (Vector Laboratories). Following 4 further washing steps the discs were postfixed for 1 h with 1% OsO4 in PBS and block contrasted with 2% uranyl acetate in H2O. After dehydration in an alcohol series (50%, 70%, 96%, 2 x 100% 15 min each) the discs were embedded in Epon and polymerized for 24 h at 60 °C. Thin sections were cut with a Reichert Ultracut E microtome and imaged with a Gatan Orius 1000 CCD camera in a Tecnai G2 spirit transmission electron microscope (FEI, Eindhoven, The Netherlands).

4.5. Immunohistochemistry

Immunostaining of Drosophila wing imaginal discs and embryos was performed according to standard protocols. Primary antibodies used in this study were: mouse anti-Wg (4D4, DSHB, 1:1000), guinea pig anti-Sens (GP55, gift from H. Bellen, Baylor College of Medicine, Houston, 1:800), rabbit anti-cleaved Casp3 (#9961, Cell Signaling, 1:200), chicken anti-bGal (ICLlab, 1:400), mouse anti-bGal (Promega, 1:2000), mouse anti-Fz2 (12A7, DSHB, 1:20), rabbit anti-Arrow (gift from S. DiNardo, 1:15000), rabbit anti-Hh (1:500), mouse anti-Ptc (DSHB, 1:100), rabbit anti-pMad (gift from Ed Laufer, Columbia University, New York, 1:1000), rabbit anti-Wls (Port et al., 2008), (1:800), guinea pig anti-Hrs (gift from H. Bellen, Baylor College of Medicine, Houston, 1:100), rabbit anti-Vps16 (gift from H. Kramer, UT Southwestern Medical School, Dallas, 1:100), rabbit anti-GFP (1:200), rabbit anti-Ecadherin (1:100).

Pictures were taken with a Zeiss LSM710 confocal microscope and the Zen software. Images were processed using Im-ageJ and Photoshop Elements.

4.6. Quantification of endosoma! compartments and Wg protein content

ywf; tuba-YFP-Rab4/5/7/11/Sp or CyO; schlank RNAi1/ hhGal4 larvae were dissected and stained with Wg antibody. Z-stacks were taken with the confocal microscope. The images were further processed using the IMARIS software. A cubic volume was defined in control and mutant compartments and the overall Wg fluorescence was measured. To mark the endosomal compartments the Surface tool was used with automatic creation and the following settings: local Background: 0.6, Threshold: 20, enable split objects, Quality: above 10. This allowed assessment of the ratio of endosomal volumes in mutant versus control compartment of each disc analyzed (N = 7 for each genotype). To quantify the amount of Wg in the respective endosomes, the fluorescence intensity of Wg signal in the previously defined endosomal compartments was measured and again set into relation of mutant to control compartment in each disc per endosome volume.

4.7. realtime PCR

Flies carrying the schlank RNAi1, and as control yw flies, were crossed to the C765Gal4 driver at 29 °C. Male larvae were dissected in three independent experiments and RNA was extracted using the Nucelospin RNA II kit (Machery-Nagel). Following an additional DNA digestion (DNA-freeTM kit, Ambion) we used the Transcriptor High Fidelity cDNA synthesis kit

(Roche Applied Science) for cDNA synthesis. Tubulin, actin and TBP expression was used for normalization of each experiment; otherwise the experiments were not normalized to show the variability between the replicates.

Author contributions

J.P. and K.B. designed and carried out the experiments; A.K. and U.L. performed the EM analysis; G.R. conducted the genome-wide screen in which schlank was found as growth regulator. J.P, G.H. and K.B wrote the paper.

Acknowledgement

We thank S. Eaton, M. Hoch, H. Kramer, S. Luschnig for providing fly stocks; H. Bellen, S. DiNardo, E. Laufer, H. Kramer for antibodies; S. Luschnig and L. Gafner for critical comments on the manuscript. This work was supported by the European Research Council, the Swiss National Science Foundation and the Forschungskredit Universität Zürich.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.mod.2013.04.001.

REFERENCES

Acharya, U., Acharya, J.K., 2005. Enzymes of sphingolipid

metabolism in Drosophila melanogaster. Cell. Mol. Life Sci. 62, 128-142.

Bauer, R., Voelzmann, A., Breiden, B., Schepers, U., Farwanah, H., Hahn, I., Eckardt, F., Sandhoff, K., Hoch, M., 2009. Schlank, a member of the ceramide synthase family controls growth and body fat in Drosophila. EMBO J. 28, 3706-3716.

Bilic, J., Huang, Y.-L., Davidson, G., Zimmermann, T., Cruciat, C.-M., Bienz, M., Niehrs, C., 2007. Wnt induces LRP6 signalosomes and promotes dishevelled-dependent LRP6 phosphorylation. Science 316, 1619-1622.

Blitzer, J.T., Nusse, R., 2006. A critical role for endocytosis in Wnt signaling. BMC Cell Biol. 7, 28.

Buechling, T., Bartscherer, K., Ohkawara, B., Chaudhary, V., Spirohn, K., Niehrs, C., Boutros, M., 2010. Wnt/Frizzled signaling requires dPRR, the Drosophila homolog of the prorenin receptor. Curr. Biol. 20, 1263-1268.

Clevers, H., Nusse, R., 2012. Wnt/ß; -catenin signaling and disease. Cell 149, 1192-1205.

Cruciat, C.-M., Ohkawara, B., Acebron, S.P., Karaulanov, E., Reinhard, C., Ingelfinger, D., Boutros, M., Niehrs, C., 2010. Requirement of prorenin receptor and vacuolar H+-ATPase-mediated acidification for Wnt signaling. Science 327, 459-463.

Dubois, L., Lecourtois, M., Alexandre, C., Hirst, E., Vincent, J.P., 2001. Regulated endocytic routing modulates wingless signaling in Drosophila embryos. Cell 105, 613-624.

Franch-Marro, X., Wendler, F., Griffith, J., Maurice, M.M., Vincent, J.-P., 2008. In vivo role of lipid adducts on Wingless. J. Cell Sci., 1587-1592.

Futerman, A.H., Riezman, H., 2005. The ins and outs of sphingolipid synthesis. Trends Cell Biol. 15, 312-318.

Hamel, S., Fantini, J., Schweisguth, F., 2010. Notch ligand activity is modulated by glycosphingolipid membrane composition in Drosophila melanogaster. J. Cell Biol. 188, 581-594.

Holopainen, J.M., Angelova, M.I., Kinnunen, P.K., 2000. Vectorial budding of vesicles by asymmetrical enzymatic formation of ceramide in giant liposomes. Biophys. J. 78, 830-838.

Kolesnick, R.N., Gorii, F.M., Alonso, A., 2000.

Compartmentalization of ceramide signaling: physical foundations and biological effects. J. Cell Physiol. 184, 285-300.

Kolesnick, R.N., Kronke, M., 1998. Regulation of ceramide production and apoptosis. Annu. Rev. Physiol. 60, 643-665.

Kraut, R., 2011. Roles of sphingolipids in Drosophila development and disease. J. Neurochem. 116, 764-778.

Le Roy, C., Wrana, J.L., 2005. Clathrin- and non-clathrin-mediated endocytic regulation of cell signalling. Nat. Rev. Mol. Cell Biol. 6, 112-126.

Lingwood, D., Simons, K., 2010. Lipid rafts as a membrane-organizing principle. Science 327, 46-50.

Lippincott-Schwartz, J., Phair, R.D., 2010. Lipids and cholesterol as regulators of traffic in the endomembrane system. Annu. Rev. Biophys. 39, 559-578.

Port, F., Kuster, M., Herr, P., Furger, E., Banziger, C., Hausmann, G., Basler, K., 2008. Wingless secretion promotes and requires retromer-dependent cycling of Wntless. Nat. Cell Biol. 10,178185.

Purvanov, V., Koval, A., Katanaev, V.L., 2010. A direct and

functional interaction between Go and Rab5 during G proteincoupled receptor signaling. Sci. Signal. 3. ra65.

Seto, E.S., Bellen, H.J., 2006. Internalization is required for proper Wingless signaling in Drosophila melanogaster. J. Cell Biol. 173, 95-106.

Silva, L.C., Ben David, O., Pewzner-Jung, Y., Laviad, E.L., Stiban, J., Bandyopadhyay, S., Merrill, A.H., Prieto, M., Futerman, A.H., 2012. Ablation of ceramide synthase 2 strongly affects biophysical properties of membranes. J. Lipid Res. 53, 430-436.

Simons, K., Ikonen, E., 1997. Functional rafts in cell membranes. Nature 387, 569-572.

Simons, K., Toomre, D., 2000. Lipid rafts and signal transduction. Nat. Rev. Mol. Cell Biol. 1, 31-39.

Singh, A., Irvine, K.D., Pontier, S.M., Schweisguth, F., 2011. Glycosphingolipids in signaling and development: from liposomes to model organisms. Dev. Dyn. 241, 92-106.

Taelman, V.F., Dobrowolski, R., Plouhinec, J.-L., Fuentealba, L.C., Vorwald, P.P., Gumper, I., Sabatini, D.D., De Robertis, E.M., 2010. Wnt signaling requires sequestration of glycogen synthase kinase 3 inside multivesicular endosomes. Cell 143,1136-1148.

Van Blitterswijk, W.J., Van Der Luit, A.H., Veldman, R.J., Verheij, M., Borst, J., 2003. Ceramide: second messenger or modulator of membrane structure and dynamics? Biochem. J. 369, 199.

Yamamoto, H., Komekado, H., Kikuchi, A., 2006. Caveolin is necessary for Wnt-3a-dependent internalization of LRP6 and accumulation of p-catenin. Dev. Cell 11, 213-223.

Zha, X., Pierini, L.M., Leopold, P.L., Skiba, P.J., Tabas, I., Maxfield, F.R., 1998. Sphingomyelinase treatment induces ATP-independent endocytosis. J. Cell Biol. 140, 39-47.