Cellular Physiology and Biochemistr and Biochemistry

347

Original Paper

Cell Physiol Biochem 2007;20:347-356 Accepted: May 03, 2007Cellular PhysiologyCellular PhysiologyCellular PhysiologyCellular PhysiologyCellular Physiologyand Biochemistrand Biochemistrand Biochemistrand Biochemistrand Biochemistryyyyy

Copyright © 2007 S. Karger AG, Basel

Fax +41 61 306 12 34E-Mail [email protected]

© 2007 S. Karger AG, Basel1015-8987/07/0205-0347$23.50/0

Accessible online at:www.karger.com/cpb

FcγγγγγRII Activation Induces Cell Surface CeramideProduction which Participates in the Assembly ofthe Receptor Signaling Complex

Marek Korzeniowski1, Abo Bakr Abdel Shakor1, AgnieszkaMakowska1, Anna Drzewiecka1, Alicja Bielawska2, KatarzynaKwiatkowska1 and Andrzej Sobota1

1The Nencki Institute of Experimental Biology, Department of Cell Biology, Warsaw, 2Departments ofBiochemistry & Molecular Biology, Medical University of South Carolina

Dr. Andrzej SobotaThe Nencki Institute of Experimental Biology02-093 Warsaw (Poland)Tel. +48-22 5892 234, Fax +48-22 822 53 42E-Mail [email protected]

Key WordsCeramide • Fc receptors • Microdomains •Sphingomyelinase • Tyrosine phosphorylation

AbstractWe studied an involvement of various cellularceramide pools in signaling of immunoreceptor FcγII(FcγRII). The cell surface ceramide level wasassessed by a technique based on binding ofceramide probes to intact cells. Total cellular ceramidewas estimated by radioactive measurements. Theactivity of sphingomyelinases was measured by NBD-ceramide release while immunoprecipitation andimmunoblotting were applied to analyze proteintyrosine phosphorylation. A complex pattern of proteinphosphorylation was found to accompany FcγRIIactivation and the phosphorylation was eitherdiminished by imipramine or increased by B13,modulators of acid sphingomyelinase and acidceramidase activity. The effects of the drugs on thephosphorylation of FcγRII and NTAL were prominent

and correlated with a reduction of the cell surfaceceramide production by imipramine and anaugmentation of the ceramide generation by B13.The ceramide generation followed activation of acidsphingomyelinase and preceded that of neutralsphingomyelinase. The level of cell surface ceramidewas additionally elevated by exogenous bacterialsphingomyelinase, but only at later stages of thereceptor activation. The total mass of ceramide wasdiminished in the course of receptor activationpointing to an engagement of enzymes metabolizingceramide. The data indicate that FcγRII activatesenzymes of the sphingomyelin cycle which affectvarious sphingomyelin/ceramide pools in a cell.

Introduction

Fcγ receptors bind IgG and initiate a wide array ofimmunoresponses, including phagocytosis of IgG-opsonized pathogens [1]. Transduction of signals by

348

FcγRII is initiated by phosphorylation of tyrosine residuesof the signaling motif located in the cytoplasmic part ofthe receptor [2]. Our previous data demonstrated thatFcγRII phosphorylation requires receptor clustering andits association with plasma membrane microdomains,known as rafts [3]. Rafts are enriched in cholesterol andsphingolipids, including sphingomyelin, and create anenvironment that facilitates a local accumulation of dual-acylated tyrosine kinases of the Src family [4]. Lyn, oneof the Src family members, is a likely candidate kinasecatalyzing the FcγRII phosphorylation which triggers thedownstream signaling cascades [3, 5]. However, themechanisms which control the association of FcγRII withrafts for signaling complex assembly are poorly known.

The enrichment of sphingomyelin in the outer leafletof the plasma membrane [6] turned our attention to theso-called sphingomyelin cycle [7] as a possible factorplaying a role in FcγRII signaling. The cycle begins withsphingomyelin hydrolysis by acid or neutralsphingomyelinases, yielding ceramide [6, 8]. Neutralsphingomyelinase (NSMase) is associated mainly withthe plasma membrane while acid sphingomyelinase(ASMase) is attributed to late endosomes and lysosomes.However, it was found recently that ASMase could bealso extruded to the cell surface and generate exofacialceramide upon activation of TNF receptor familymembers and UV-irradiation of cells. The ceramideformed was required for efficient clustering of deathreceptors adding a new aspect to pro-apoptotic functionsof ceramide [9-11]. Aside from the role of ceramide ininduction of apoptosis, the lipid was recently shown to beessential for clustering of immunoreceptor FcγRII in raftsand, hence, for FcγRII signal generation [12]. In this reportwe continue our studies on FcγRII-triggered ceramideproduction and analyze how FcγRII clustering affectsthe level of various cellular ceramide pools by activationof ASMase and NSMase.

Materials and Methods

Cell culture and FcγRII activationU937 cells were cultured as described [3]. For FcγRIIA

cross-linking, cells were incubated with mouse anti-FcγRIIantibody (IgG2b clone IV.3, 30min, 4°C) followed by goat anti-mouse F(ab)2 IgG (1-20min, 4°C; ICN). The stage immediatelyafter incubation of the cells with anti-FcγRII only, which doesnot induce receptor activation [3,13], was set as time point “0”of receptor cross-linking. Imipramine at 50-70 µM or N,N-dimethylformamide up to 0.03% as imipramine carrier (bothSigma) were applied for 1h (37°C) prior to incubation of the

cells with anti-FcRII. Bacterial sphingomyelinase (bSMase;Sigma) was used at 40-70 mU/ml for 4min at 4°C. B13 at 10 µMand 0.05% dimethyl sulfoxide, a B13 carrier, were used overnightat 37°C in the growth medium.

Ceramide determinationCell surface ceramide: cells (3x106/sample) were fixed with

1% formaldehyde and incubated with 1 µg/ml anti-ceramideIgM (clone MID 15B4, Alexis; 45min, 20°C). Cell-boundantibody was desorbed for 30 s with ice-cold 100 mM glycine-HCl, pH 3.0, neutralized, spotted onto nitrocellulose and revealedwith goat anti-mouse IgM conjugated with peroxidase.Immunoreactive spots were visualized by chemiluminescence(Pierce) and quantified densitometrically in a Fluor MultiImagerusing Quantity One software (BioRad) [11]. Alternatively,exofacial ceramide level was analyzed using FACScan and CellQuest software (Becton Dickinson) after incubation of the cellswith anti-ceramide and anti-mouse IgM goat IgG-FITC (JacksonImmunoResearch).

Total cellular level of ceramide: [3H]serine (1.5 µCi/ml,Perkin Elmer) was administrated to the cells (1x106/ml) for 48hin RPMI containing 6% dialyzed FBS. Lipids were extracted(1x107cell/sample) [14], subjected to alkaline hydrolysis [15]and analyzed by thin layer chromatography on silica gel 60 F254using chloroform/acetic acid (9:1) mixture. Ceramide wasidentified by iodine vapor staining, in reference to a ceramidestandard (Sigma), scraped off and quantified by liquidscintillation counting in a Beckman counter.

ASMase and NSMase activity assaysFor ASMase activity assay, cells (3x106/sample) were lysed

in 120 µl of 0.5% Triton X-100, 50 mM acetate buffer, pH 5.0.After 20min, lysates were centrifuged (14000g, 5min, 4°C) and100 µl of supernatant was mixed with 150 µl of 2 mM EDTA,200 mM acetate buffer, pH 5.0, containing liposomes composedof 12 µM NBD-sphingomyelin (Molecular Probes), 24 µMsphingomyelin, 24 µM cholesterol and 108 µM 1,2-dioleoyl-sn-glycerol-3-phosphocholine (all from Sigma). The liposomes wereprepared by sonication of lipids suspended in 70 mM NaCl.For determination of NSMase activity, cells were lysed in 0.5%Triton X-100 in the presence of 10 mM MgCl2, 2 mM EDTA,1 mM ATP, 30mM p-nitrophenylphosphate, 0.2 mM Na3VO4,10 mM β-glycerophosphate, 5 mM DTT, 20 mM Hepes, pH 7.4,and protease inhibitor cocktail [16]. After preclearing, 100 µl oflysate was mixed with 150 µl of 100 mM NaCl, 2 mM MgCl2,50 mM Hepes, pH 7.4, containing liposomes as described above.After 1h at 37°C, NBD-ceramide was isolated [17] and measuredin a fluorimeter (460 nm /534nm).

Electron microscopyPlasma membrane sheets were obtained by mechanical

cleavage of cells [18]. The cytoplasmic surface of the plasmamembrane was probed for 2h at 20°C with rabbit anti-FcγRIIantibody and goat or donkey anti-rabbit IgG-gold (10 nm gold;Sigma, Aurion). In a series of experiments, prior to cell cleavagethe cells were incubated with 4 µg/ml probe containing thecysteine-rich domain of kinase suppressor of RAS fused withglutathione S-transferase (GST) which specifically recognizes

Korzeniowski/Shakor/Makowska/Drzewiecka/Bielawska/Kwiatkowska/Sobota

Cell Physiol Biochem 2007;20:347-356

349

ceramide, as shown previously [11]. The probe was followedby goat anti-GST (Rockland) and donkey anti-goat IgG-gold(6 nm, Aurion). Samples were fixed with 2.5% glutaraldehyde in50 mM cacodylate buffer (15min, 20°C), postfixed with 1% OsO4in the buffer (10min, 20°C) and counterstained with 1% uranylacetate in water. The samples were examined under a JEM-1200EX (JEOL) electron microscope. Gold particles marking ceramidewere scored from micrographs (2 µm2 of the membrane surface).The calculations were performed for two independent

experiments; at least 10 micrographs were analyzed for eachset of experiments.

SDS-PAGE and immunoblottigTo obtain whole cell lysates, cells (5x105/sample) were

lysed in ice-cold 0.2% Triton X-100, 150 mM NaCl, 2 mM EDTA,2 mM EGTA, 30 mM Hepes, pH 7.4, with protease andphosphatase inhibitors. After 10min, SDS-sample buffer wasadded, the samples were boiled and proteins were separated

Fig. 1. FcγRII cross-linking induces protein tyrosine phosphorylation that is sensitive to modulators of ASMase and acidceramidase activity. (A, B) Time-course of protein tyrosine phosphorylation during FcγRII cross-linking (4°C) in U937 cellspreincubated without and with 50 µM imipramine (A) and 10 µM B13 (B). Molecular mass standards are shown in kDa on the left.(C-D) Densitometric quantification of phosphoproteins revealed in (A and B) without (closed symbols) and with drug pretreatment(open symbols) after normalization against actin content. The phosphorylation of a particular band is expressed in relation to itsmaximum value in control cells non-exposed to the drugs, that was arbitrarily set at 100. (E, F) Tyrosine phosphorylation ofFcγRIIA (doublet) and NTAL revealed after immunoprecipitation of the proteins from whole cell lysates. Prior to FcγRII cross-linking, cells were preincubated with 50 µM imipramine or 10 µM B13. The immunoprecipitates were probed for the presence oftyrosine phosphoproteins (upper panels) and either biotin-labeled IV.3 anti-FcγRII reflecting the receptor content or NTAL (lowerpanels). Results of one of three experiments are shown.

Cell Surface Ceramide in FcγRII Signaling Cell Physiol Biochem 2007;20:347-356

350

by 10% SDS-PAGE. After transfer to nitrocellulose sheets, theywere probed with mouse anti-phosphotyrosine, clone PY66(Santa Cruz Biotechnology) and goat anti-mouse IgG-peroxidase (ICN). The membranes were reprobed, after stripping,with mouse anti-actin IgG (Roche). Immunoreactive bands werevisualized with the chemiluminescent substrate and quantifieddensitomerically as described above.

Immunoprecipitation of FcγRII was performed from celllysates obtained in the presence of 0.2% Triton X-100 and1.5% N-octylglucoside essentially as described earlier [3]. Inthis set of experiments, biotin-labeled IV.3 and rabbit anti-mouseIgG (Jackson ImmunoResearch) were used for FcγRII cross-linking. For immunoprecipitation of NTAL, rabbit-anti-NTALwas applied. The precipitates were probed with mouse anti-phosphotyrosine and, in parallel, either with goat anti-biotinperoxidase conjugated IgG (Sigma) or with mouse anti-NTALantibodies to reveal amounts of the precipitated receptor andNTAL.

Data were evaluated by one-way ANOVA test with p<0.05considered significant.

Results

FcγRII cross-linking induces activation ofASMase, NSMase and asynchronous proteinphosphorylationFcγRII of U937 monocytes was activated by cross-

linking with mouse anti-FcγRII antibody followed by anti-mouse IgG. The receptor clustering induced tyrosinephosphorylation of several proteins (Fig. 1A, B), althougha detailed densitometric analysis revealed that the time-course of the phosphorylation of individual bands varied.A 130 kDa band and a prominent 120 kDa band werephosphorylated progressively during receptor cross-linkingin an approximately linear fashion, while a 65 kDa bandunderwent abrupt phosphorylation reaching the maximalphosphorylation level after about 5 min of the receptorclustering (Fig. 1C, D, closed circles). On the other hand,a 30 kDa band, identified as an adaptor protein NTAL byimmunoprecipitation (see below), displayed a significantlydelayed tyrosine phosphorylation which started at 5 minof the receptor cross-linking (Fig. 1A).

The level of the phosphorylation was diminished tovarious extents in cells pretreated with imipramine (Figs.1A, and 1C, open circles), an ASMase inhibitor withrecently discovered activity against acid ceramidase aswell [19,20]. Notably, imipramine reduced tyrosinephosphorylation of FcγRII and affected alsophosphorylation of NTAL, as found afterimmunoprecipitation of the proteins (Fig. 1E, F). Anestimation of sphingomyelinase activity in whole celllysates revealed that FcγRII cross-linking induced fast

and transient elevation of ASMase activity with themaximum at 5min of the clustering. In contrast, theactivity of NSMase rose progressively, peaking at a laterstage of the receptor cross-linking, at 11min (Fig. 2). Asexpected, pretreatment of cells with imipramine stronglyinhibited the activity of ASMase in derived cell lysates(Fig 2, compare open and closed circles). On the contrary,imipramine slightly activated (by 15-20%) rather thaninhibited the activity of bSMase in vitro (not shown). Thisis in agreement with earlier data that the drug exerts itsinhibitory effect on ASMase in intact cells by inductionof proteolytic degradation of the enzyme in lysosomes[19].

To further explore the involvement of ceramide inFcγRII-induced protein phosphorylation, the cells wereexposed to B13, a potent blocker of acid ceramidase [21],a critical scavenging enzyme of ceramide. An analysis ofoverall protein tyrosine phosphorylation revealed that, incontrast to the imipramine action, pretreatment of cellswith B13 accelerated and augmented the FcγRII-inducedphosphorylation. The reinforcing effect of B13 wasclearly manifested at early stages of FcγRII cross-linking,at 5-10min, as seen for a 65 kDa and a 75 kDaphosphoproteins (Fig. 1B, D). The data were supported

Fig. 2. FcγRII cross-linking activates ASMase and NSMasedifferently. Activity of ASMase (circles) and NSMase (squares)in whole lysates of U937 cells during FcγRII cross-linking.Open circles indicate activity of ASMase in cells pretreatedwith 70 µM imipramine prior to FcγRII activation. The enzymeactivities are expressed in relation to their maximum valuesequalized to 100. The data shown are mean ± SEM from fourexperiments.



351

Fig. 3. FcγRII cross-linking induces cellsurface ceramide production. (A, B)Ceramide detected on the surface of intactU937 cells by anti-ceramide IgM bindingand dot-blotting assay during FcγRIIAcross-linking (A) and after incubation ofcells with 40-70 mU/ml bSMase (4min,4°C) without receptor cross-linking (B).The upper panels show representativedot-blots, while lower panels displayprofiles of ceramide generation. (C, D).Quantification of cell surface ceramideproduction as a function of concentrationof anti-FcγRII. The quantification isbased on dot-blotting (C) and FACScananalysis (D) of cell-surface bound anti-ceramide IgM. Anti-FcγRII was followedby 10 µg/ml goat anti-mouse F(ab)2. (E,F) Cell surface ceramide production incells treated with 70 µM imipramine (E)and 10 µM B13 (F) prior to FcγRIIA cross-linking. The data in (A, C) are the mean ±SEM from two-three experiments, in (B,D-F) results of representative experimentsare shown.

by examining phosphorylation of immunoprecipitatedFcγRII and NTAL. Under the influence of B13 bothproteins demonstrated a marked increase ofphosphorylation at early steps of FcγRII clustering andthe phosphorylation of NTAL sustained at the elevatedlevel later on (Fig 1E, F).

FcγRII cross-linking triggers generation of cellsurface ceramide and affects the level of totalcellular ceramideTo examine whether the FcγRII-induced protein

phosphorylation correlated with ceramide production inthe cells we first evaluated the level of cell surfaceceramide in an assay based on immunodetection of ananti-ceramide antibody bound to intact cells. This approachrevealed that the exofacial ceramide level increasedtransiently by about 3-fold between 5-10min of FcγRII

cross-linking (Fig. 3A). The adequacy of the ceramidedetection by the anti-ceramide-binding procedure wasconfirmed by treatment of the cells with exogenousbSMase, without FcγRII cross-linking. Under theseconditions, the amounts of ceramide detected on the cellsurface were proportional to the amounts of theexogenous bSMase applied (Fig. 3B). A transientelevation of exofacial ceramide during FcγRII cross-linking was also detected when surface-bound anti-ceramide was analyzed by FACScan. However, themagnitude of the changes was lower due to the highbackground of fluorescence found in cells prior to FcγRIIcross-linking (see Fig. 4A).

We measured the exofacial ceramide generation asa function of the amount of activated FcγRII. Dot-blottingand FACScan analysis yielded very consistent resultsshowing that the ceramide production was triggered


352

sharply by low doses of anti-FcγRII used for receptorcross-linking and reached a plateau at 0.5-1 µg/ml of theantibody (Fig. 3C, D). Therefore, in the course of thework, 3 µg/ml of anti-FcγRII was used routinely to ensuremaximal receptor activation.

The level of cell surface ceramide produced duringFcγRII activation was altered in cells pretreated withimipramine and B13, the modulators of ASMase and acidceramidase activity. Imipramine reduced the amount ofthe exofacial ceramide while B13-treated cells had

Fig. 4. Exogenous bSMase intensifies cell surface ceramideproduction at later stages of FcγRII cross-linking. (A, B)FcγRIIwas cross-linked in U937 cells for 7 and 12min (4°C) andduring the last 4min of the procedure 70 mU/ml bSMase wasapplied. Cell surface ceramide was estimated by binding ofanti-ceramide IgM and FACScan analysis (A), while the totalcellular ceramide level was analyzed by thin layerchromatography (B). The data in (A, B) are the mean ± SEM ofthree experiments. Differences significant at p<0.05 are markedby bars, differences statistically insignificant are not marked.(C) Dot blotting analysis of cell surface ceramide productionduring FcγRII cross-linking with and without bSMase.

significantly increased cell surface ceramide level (Fig.3E, F). The increase was detectable at 5min of FcγRIIcross-linking and the amount of the ceramide in B13-exposed cells sustained elevated for 20min compared witha transient ceramide accumulation in the control (Fig. 3E,F). The opposed influence of imipramine and B13 on cellsurface ceramide generation during FcγRII activationcorrelated well with the inhibition and the augmentationof FcγRII-induced protein tyrosine phosphorylation seenin cells exposed to the drugs (see Fig. 1).

Having established that FcγRII cross-linking iscorrelated with a transient rise of cell surface ceramide,we examined whether these changes were reflected bymodulation of the total cellular ceramide level estimatedby radioactive measurements of [3H]ceramide. It wasfound that the total mass of ceramide in the cells wasprogressively diminishing in the course of FcγRII cross-linking. After 12min of the receptor activation, the totalceramide level fell to 63% of its value found in the cellsbefore receptor activation (Fig 4B).

Bacterial SMase augments ceramide productiononly at later stages of FcγRII cross-linkingSince the exofacial ceramide is generated from

sphingomyelin most likely residing in the outer leaflet ofthe plasma membrane, we assumed that treatment of cellswith bSMase could augment this process. FACScananalysis revealed that bSMase applied for 4min withoutFcγRII cross-linking elevated the cell surface ceramidelevel by 1.4-fold. For comparison, receptor cross-linkingled to a 1.8-fold and 1.6-fold increase of the exofacialceramide level after 7 and 12 min, respectively (Fig. 4A).It is of note that treatment of the cells with bSMase forthe last 4min in the course of 7 minute-long receptor cross-linking did not affect the level of surface ceramide (Fig4A). However, application of the exogenous enzymeduring the last 4 min of receptor cross-linking lasting fora total of 12 minutes markedly elevated the amount ofcell surface ceramide. The combined treatment resultedin a 1.7-fold increase of the exofacial ceramide level abovethat found after FcγRII cross-linking alone (Fig. 4A). TheFACScan analysis was supported by dot-blotting whichshowed an intensification of ceramide production whenbSMase was added for 4min and followed by 15-20minof FcγRII cross-linking (Fig. 4C). This additive effect ofFcγRII activation and bSMase action on the ceramidelevel was also detectable when the level of total cellularceramide was estimated. bSMase applied for 4min at theend of a 12min period of receptor cross-linking doubledthe amount of the total ceramide (Fig. 4B).



353

Fig. 5. Distribution of FcγRII and ceramide in the plasma membrane during receptor cross-linking. (A) FcγRII in cells exposed toanti-FcγRII only. (B-C) Formation of electron-dense structures accommodating FcγRII after 5min (B) and 20min (C) of receptorcross-linking. (D) Colocalization of FcγRII (10 nm gold) and cell surface ceramide (6 nm gold) after 5min of FcγRII cross-linking.(E) Quantitation of gold particles reflecting ceramide on the external surface of the plasma membrane in cells exposed to anti-FcγRII only and after 5 min of FcγRII cross-linking. The data are the mean ± SEM. from two experiments. Bars, 100 nm.

Cell surface ceramide associates with clustersof activated FcγRIIFor ultrastructural analysis of the distribution of cell

surface ceramide during FcγRII cross-linking we usedlarge plasma membrane sheets obtained by mechanicalcleavage of cells. As seen in Fig. 5A, the receptor wasuniformly dispersed in the plane of the plasma membraneof unstimulated cells. Upon FcγRII cross-linking, electron-dense structures were formed in the membrane; thestructures were distinct at 5 min and their size increasedwith time of receptor activation. These clusters

accommodated the vast majority of the activated FcγRII(Fig. 5B, C).

When the external surface of the plasma membranewas examined for ceramide presence, prominent labelingof the lipid was found at FcγRII clusters after 5 min ofreceptor cross-linking (Fig. 5D). Quantitation of goldparticles attributed to the cell surface ceramidedemonstrated that the vast majority of the exofacialceramide was associated with the electron-densestructures, whereas minute amounts of the lipid weredetected outside the structures. Only traces of ceramide


354

were seen on the surface of resting cells (Fig. 5E).

Discussion

It has been established that activation of FcγRIItriggers signaling cascades which require protein tyrosinephosphorylation [13, 22, 23]. Our present analysisdemonstrates differences in the time course of thephosphorylation concerning various proteins suggestingthat these proteins may be involved at different stages ofFcγRII signaling. Electron microscopy studies revealedthat the receptor signaling was accompanied by formationof electron-dense structures in the plasma membraneaccommodating FcγRII. Since FcγRII is known toassociate with rafts for phosphorylation [3, 11, 22], andsince Lyn kinase, CD55 and phosphoproteins have beenshown to occupy the electron-dense structures [18], theformation and progressive enlargement of the structuresis likely to reflect raft coalescence during FcγRII signalingcomplex formation. The electron-dense structures werefound to accommodate cell surface ceramide (Fig. 5D,E).

An immunochemical analysis of the exofacialceramide revealed that the ceramide level was transientlyelevated upon FcγRII activation. We found that two variousmodulators of the sphingomyelin cycle affected FcγRII-induced ceramide production in an opposite manner.Imipramine diminished the level of the ceramide andconcomitantly inhibited tyrosine phosphorylation of proteinsaccompanying FcγRII activation. On the contrary, B13accelerated and augmented accumulation of ceramideon the cell surface during FcγRII cross-linking.Simultaneously, under these conditions the protein tyrosinephosphorylation underwent also rapid elevation in amanner reflecting the profile of exofacial ceramideappearance. We obtained similar data (not shown) withan use of LCL204, a new B13 analogue considered asan acid ceramidase inhibitor [24]. The exofacial ceramideseems to affect strongly the phosphorylation of FcγRIIand NTAL, judging from the influence of imipramine andB13 on the process. This may suggest that transmembraneproteins are especially sensitive to the presence ofceramide in the outer leaflet of the plasma membrane.The present data are in line with our earlier observationsthat administration of exogenous C16-ceramide facilitatedclustering of cross-linked FcγRII and an association ofthe receptor with rafts [12]. Taken together, the dataindicate that once generated upon FcγRII activation, theceramide can facilitate translocation of proteins in the

plane of the plasma membrane and govern coalescenceof rafts into FcγRII signaling platforms where proteinphosphorylation occurs. These processes are modulatedby ceramide due to tendency of the lipid to self-aggregateand laterally separate into domains [25].

Imipramine and closely related desipramine wereconsidered for a long time as drugs inducing proteolyticdegradation of ASMase [19]. However, it wasdemonstrated recently that acid ceramidase is alsosusceptible to the proteolysis [20]. Without further analysisit is hard to reconcile whether effect of imipramine oncells can be ascribed either to lack of ceramide generation(ASMase inhibition) or accumulation of ceramide and lackof its derivatives, such as sphingosine and sphingosine-1-phosphate (inhibition of acid ceramidase). Indeed, wefound that pretreatment of cells with imipramine let to70% inhibition of ASMase (Fig. 2) but also diminishedactivity of acid ceramidase by nearly 50% (not shown).It is of note that, regardless of this potentially dual actionof imipramine, in our hands the drug reduced both thelevel of cell surface ceramide and protein tyrosinephosphorylation during FcγRII cross-linking, as oppositeto the B13 action. Therefore, we assume that theassembly of FcγRII signaling complex requires, at leastat the onset, the presence of cell surface ceramidegenerated presumably by ASMase, the target ofimipramine. In agreement with this assumption, the peakof the ceramide generation during FcγRII cross-linkingin cells not exposed to the drugs followed activation ofASMase and preceded activation of NSMase. Thesupposition that imipramine restrained the assembly ofthe FcγRII signaling complex by inhibition of ASMaseactivity is supported by our earlier findings that exogenousC16-ceramide partially reversed the inhibitory effect ofimipramine exerted on FcγRII clustering andphosphorylation [12].

An involvement of NSMase in FcγRII-inducedprotein phosphorylation might be also considered,especially in view of the data that imipramine can augmentthe activity of this enzyme, while inhibiting ASMase [26].Previously, ceramide was suggested to regulate the levelof protein phosphorylation by activation of a kinase and aphosphatase, both specific toward serine/threonineresidues. Production of intracellular ceramide controllingthese enzymes was attributed to the activity of plasma-membrane-associated NSMase [8]. A joint participationof ASMase and NSMase in the signaling cascadestriggered by TNF in U937 cells was demonstrated.However, the two enzymes controlled distinct, non-overlapping pathways of the TNF receptor signaling [16].



355

The findings presented here indicate that ASMase isresponsible for production of the cell surface ceramideduring FcγRII clustering, while the role of the intracellularceramide generated by NSMase at later stages of thereceptor activation remains unknown. Yet another sourceof cell surface ceramide can be hydrolysis of glycolipids,although participation of this process in signaling pathwaysremains obscure [27].

The exofacial ceramide is presumably not thepredominant fraction of total cellular ceramide sincedespite a temporal increase of its level, the total mass ofthe lipid was diminished in the cells (Fig. 4). Moreover,these data indicate that FcγRII cross-linking triggersactivation of enzymes of the sphingomyelin cycle whichgenerate the exofacial ceramide and simultaneouslymetabolize the intracellular pool of the lipid. Thissupposition is in line with earlier suggestions that themembrane sphingomyelin cycle imposes a quantitativecontrol on the interplay between T-cell receptor and deathreceptors affecting eventually T cell activation andsurvival [28].

The level of ceramide was synergistically elevatedby bSMase and FcγRII cross-linking at later stages ofreceptor activation. This suggests that at the beginningof FcγRII clustering, exogenous bSMase and endogenousreceptor-activated ASMase can target a common poolof exofacial sphingomyelin. As a result, at the initial stageof FcγRII activation, the effect of these enzymes onceramide production is not additive. Later on, two separatepools of sphingomyelin become available for the enzymesand one can assume that endogenous ASMase is coupledto the raft-based signaling complexes of FcγRII whilebSMase utilizes sphingomyelin exposed/synthesized in theplasma membrane during assembly of the complex. Thissphingomyelin fraction located outside the receptor

signaling platforms is likely to acquire a fluid state thathas been found as a requirement for sphingomyelinhydrolysis by bSMase [29]. The existence of distinctsphingomyelin pools in the plasma membrane has beensuggested before. However, the “signaling” pool of thelipid utilized upon prolonged activation of TNFα receptorand bSMAse treatment was ascribed to sphingomyelinlocated in the inner leaflet of the plasma membrane [30],in contrast to the results of the present work implicatingthe outer leaflet.

Abbreviations

ASMase (acid sphingomyelinase); bSMAse(bacterial sphingomyelinase);GST (glutathioneS-transferase); NSMase (neutral sphingomyelinase).

Acknowledgements

We thank dr. V. Horejsi of Institute of MolecularGenetics, Prague, Czech Republik for rabbit and mouseanti-NTAL antibodies, dr. E. Gulbins (University of Essen,Germany) for the ceramide probe, dr. J.-L. Theillaud(Unite INSERM 255, Centre de Recherches Biomedicalesdes Cordeliers, France) for rabbit anti-FcγRII serum andLipidomic Core, Medical University of South Carolina,USA, for providing B13. We also thank dr. AgnieszkaStrzelecka-Kiliszek for initial experiments with electronmicroscopy and Kazimiera Mrozinska for excellenttechnical assistance. This work was supported by grantfrom the Polish Ministry of Education and Science no.2P04C 141 29.

References

1 Swanson JA, Hoppe AD: Thecoordination of signaling during Fcreceptor-mediated phagocytosis. JLeukoc Biol 2004;76:1093-1103.

2 Mitchell MA, Huang MM, Chien P, IndikZK, Pan XQ, Schreiber AD: Substitutionsand deletions in the cytoplasmic domainof the phagocytic receptor FcγRIIA:Effect on receptor tyrosinephosphorylation and phagocytosis.Blood 1994;84:1753-1759.

3 Kwiatkowska K, Frey J, Sobota, A:Phosphorylation of FcγRIIA is requiredfor the receptor-induced actinrearrangement and capping: the role ofmembrane rafts. J Cell Sci 2003;116:537-550.


356

4 Simons K, Toomre D: Lipid rafts andsignal transduction. Nature Rev Mol CellBiol 2000;1:31-39.

5 Bewarder N, Weinrich V, Budde P,Hartmann D, Flaswinkel H, Reth M, FreyJ: In vivo and in vitro specificity ofprotein tyrosine kinases forimmunoglobulin G receptor (FcγRII)phosphorylation. Mol Cell Biol1996;16:4735-4743.

6 Kolesnick RN, Goni FM, Alonso A:Compartmentalization of ceramidesignaling: physical foundations andbiological effects. J Cell Physiol2000;184:285-300.

7 Futerman AH, Hannun YA: The complexlife of simple sphingolipids. EMBO Rep2004;5:777-782.

8 Marchesini N, Hannun YA: Acid andneutral sphingomyelinases: roles andmechanisms of regulation. Biochem. CellBiol 2004;82:27-44.

9 Grassme H, Jekle A, Riehle A, SchwarzH, Berger J, Sandhoff K, Kolesnick R,Gulbins E: CD95 signaling via ceramide-rich membrane rafts. J Biol Chem2001;276:20589-20596.

10 Grassme H, Jendrossek V, Bock, J, RiehleA, Gulbins E: Ceramide-rich membranerafts mediate CD40 clustering. J Immunol2002;168:298-307.

11 Charruyer A, Grazide S, Bezombes C,Muller S, Laurent G, Jaffrezou JP: UV-Clight induces raft-associated acidsphingomyelinase and JNK activationand translocation independently on anuclear signal. J Biol Chem2005;280:19196-19204.

12 Abdel Shakor AB, Kwiatkowska K,Sobota A: Cell surface ceramidegeneration precedes and controls FcγRIIclustering and phosphorylation in rafts.J Biol Chem 2004;279:36778-36787.

13 Kwiatkowska K, Sobota A: The clusteredFcγ receptor II is recruited to Lyn-containing membrane domains andundergoes phosphorylation in acholesterol-dependent manner. Eur JImmunol 2001;31:989-998.

14 Bligh EG, Dyer WJ: A rapid method oftotal lipid extraction and purification.Can J Biochem Physiol 1959; 7:911-917.

1 5 Santana P, Fanjul LF, Ruiz de GalarretaCM: Measurement of sphingomyelin andceramide cellular levels afters p h i n g o m y e l i n a s e - m e d i a t e dsphingomyelin hydrolysis; inPhospholipid signaling protocols, IMBird (ed): Humana Press, Totowa, NJ,1998, pp 217-221.

16 Wiegmann K, Schutze S, Machleidt T,Witte D, Kronke M: Functionaldichotomy of neutral and acidicsphingomyelinases in tumor necrosisfactor signaling. Cell 1994;78:1005-1015.

17 Hinkovska-Galcheva V, Kjeldsen L,Mansfield PJ, Boxer LA, Shayman JA,Suchard SJ: Activation of a plasmamembrane-associated neutralsphingomyelinase and concomitantceramide accumulation during IgG-dependent phagocytosis in humanpolymorphonuclear leukocytes. Blood1998;91:4761-4769.

18 Strzelecka-Kiliszek A, Korzeniowski M,Kwiatkowska K, Mrozinska K, SobotaA: Activated FcγRII and signallingmolecules revealed in rafts by ultra-structural observations of plasma-membrane sheets. Mol Membr Biol2004;21:101-108.

19 Hurwitz R, Ferlinz K, Sandhoff K: Thetricyclic antidepressant desipraminecauses proteolytic degradation oflysosomal sphingomyelinase in humanfibroblasts. Biol Chem Hoppe-Seyler1994;375:447-450.

20 Zeidan YH, Pettus BJ, Elojeimy S, TahaT, Obeid LM, Kawamori T, Norris JS,Hannun YA: Acid ceramidase but not acidsphingomyelinase is required for tumornecrosis factor-α-induced PGE2

production. J Biol Chem2006;281:24695-24703.

21 R. Mailloux, Selzner M, Bielawska A,Morse MA, Rudiger HA, Sindram D,Hannun YA, Clavien PA: Induction ofapoptotic cell death and prevention oftumor growth by ceramide analogues inmetastatic human colon cancer. CancerRes 2001;61:1233-1240.

22 Katsumata O, Hara-Yokoyama M,Sautes-Fridman C, Nagatsuka Y, KatadaT, Hirabayashi Y, Shimizu K, Fujita-Yoshigaki J, Sugiya H, Furuyama S:Association of FcγRII with low-densitydetergent-resistant membranes isimportant for cross-linking-dependentinitiation of the tyrosinephosphorylation pathway and superoxidegeneration. J Immunol 2001;167:5814-5823.

23 Ganesan LP, Fang H, Marsh CB,Tridandapani S: The protein-tyrosinephosphatase SHP-1 associates with thephosphorylated immunoreceptortyrosine-based activation motif ofFcγRIIa to modulate signaling events inmyeloid cells. J Biol Chem2003;278:35710-35717.

24 Holman DH, Turner LS, El-Zawahry A,Elojeimy S, Liu X, Bielawski J, Szulc ZM,Norris K, Zeidan YH, Hannun YA,Bielawska A, Norris JS: Lysosomotropicacid ceramidase inhibitor inducesapoptosis in prostate cancer cells. CancerChemother Pharmacol 2007;in press.

25 van Blitterswijk WJ, Van Der Luit AH,Veldman RJ, Verheij M, Borst J: Ceramide:second messenger or modulator ofmembrane structure and dynamics?Biochem J 2003;369:199-211.

26 Jensen J-M, Schutze S, Forl M, KronkeM, Proksch E: Roles for tumor necrosisfactor receptor p55 and sphingomelinasein repairing the cutaneous permeabilitybarrier.J Clin Invest 1999;104:1761-1770.

27 Ogretmen B, Hannun YA: Biologicallyactive sphingolipids in cancerpathogenesis and treatment. Nat RevCancer 2004;4:604-616.

28 Gombos I, Kiss E, Detre C, Laszlo G,Matko J: Cholesterol and sphingolipidsas lipid organizers of the immune cells’plasma membrane: their impact on thefunctions of MHC molecules, effectorT-lymphocytes and T-cell death.Immunol Lett 2006;104:59-69.

29 Ruiz-Arguello MB, Veiga MP, ArrondoJL, Goni FM, Alonso A:Sphingomyelinase cleavage ofsphingomyelin in pure and mixed lipidmembranes. Influence of the physicalstate of the sphingolipid. Chem PhysLipids 2002;114:11-20.

30 Linardic CM, Hannun YA: Identificationof a distinct pool of sphingomyelininvolved in the sphingomyelin cycle. JBiol Chem 1994;269:23530-23537.



Cellular Physiology and Biochemistr and Biochemistry

Documents

Transcript of Cellular Physiology and Biochemistr and Biochemistry