Báo cáo Y học: Neurotrophic signalling pathway triggered by prosaposin in PC12 cells occurs through lipid rafts doc

Neurotrophic signalling pathway triggered by prosaposin in PC12 cells occurs through lipid rafts Maurizio Sorice 1,2 , Sabrina Molinari 1 , Luisa Di Marzio 3 , Vincenzo Mattei 1,2 , Vincenzo Tasciotti 1,2 , Laura Ciarlo 1 , Masao Hiraiwa 4 , Tina Garofalo 1,2 and Roberta Misasi 1 1 Dipartimento di Medicina Sperimentale, ‘Sapienza’ University, Rome, Italy 2 Laboratorio di Medicina Sperimentale e Patologia Ambientale, ‘Sapienza’ University, Rieti, Italy 3 Dipartimento di Scienze del Farmaco, Universita ` G. D’Annunzio, Chieti Scalo, Italy 4 Department of Neurosciences, University of California, San Diego, CA, USA Prosaposin is a neurotrophic factor that has been dem- onstrated to mediate trophic signalling events in differ- ent cell types through its active region within the saposin C domain [1,2]. Prosaposin is also secreted into various body fluids [3], and mRNA and protein levels increase following peripheral nerve injury. Exogenous prosaposin promotes axonal sprouting in neural cells and myelin lipid synthesis, and prolongs cell survival in both Schwann cells and oligodendrocytes [4], suggesting that secreted prosaposin may have neurotrophic and neuroprotective roles. Moreover, prosaposin treatment induced pheochromocytoma cells (PC12) to enter the S-phase of the cell cycle [5] and monocytic U937 cell death prevention, reducing both necrosis and apoptosis [6]. This effect was achieved through rapid extracellular signal-regulated kinase (ERK) phosphorylation, sphin- gosine kinase (SphK) activation, with intracellular sphingosine 1-phosphate (S1P) production, and phos- phatidylinositol 3-kinase–Akt pathway involvement. Thus, prosaposin appears to be a regulatory factor in the ceramide–S1P rheostat, which regulates cell fate, not only in cells of neurological origin [6]. Prosaposin triggers the signal cascade after binding to a putative G o -coupled cell surface receptor [7] and ⁄ or to the low-density lipoprotein (LDL) receptor- related protein (LRP-1) [8]. The presence of lipid-binding domains in prosaposin and saposins has been well demonstrated [9], and a Keywords lipid domains; PC12 cells; prosaposin; rafts; sphingosine kinase Correspondence R. Misasi, Department of Experimental Medicine, ‘Sapienza’ University, Viale Regina Elena 324, Rome 00161, Italy Fax: +39 (6) 4454820 Tel: +39 (6) 49970663 E-mail: roberta.misasi@uniroma1.it (Received 6 May 2008, revised 25 July 2008, accepted 5 August 2008) doi:10.1111/j.1742-4658.2008.06630.x Prosaposin is a neurotrophic factor that has been demonstrated to mediate trophic signalling events in different cell types; it distributes to surface membranes of neural cells and also exists as a secreted protein in different body fluids. Prosaposin was demonstrated to form tightly bound complexes with a variety of gangliosides, and a functional role has been suggested for ganglioside–prosaposin complexes. In this work, we provide evidence that exogenous prosaposin triggers a signal cascade after binding to its target molecules on lipid rafts of pheochromocytoma PC12 cell plasma mem- branes, as revealed by scanning confocal microscopy and linear sucrose gradient analysis. In these cells, prosaposin is able to induce extracellular signal-regulated kinase phosphorylation, sphingosine kinase activation, and consequent cell death prevention, acting through lipid rafts. These findings point to the role of lipid rafts in the prosaposin-triggered signalling path- way, thus supporting a role for this factor as a new component of the multimolecular signalling complex involved in the neurotrophic response. Abbreviations CTxB, cholera toxin B subunit; D-PDMP, D-threo-1-phenyl-2-decanoylamin-3-morpholino-propanol; ERK, extracellular signal-regulated kinase; FITC, fluorescein isothiocyanate; HRP, horseradish peroxidase; HS, horse serum; LDL, low-density lipoprotein; LRP-1, low-density lipoprotein receptor-related protein; MbCD, methyl-b-cyclodextrin; PI, prodium iodide; p-Ser, phosphoserine; PT, pertussis toxin; S1P, sphingosine 1-phosphate; SphK, sphingosine kinase. FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS 4903 role has been suggested for ganglioside–prosaposin complexes as structural components of membrane sig- nalling lipid domains [10]. In the last few years, lipid rafts have been characterized in different cell types as small and highly dynamic structures, envisaged as lat- eral assemblies of specific lipids and proteins in cellular membranes, proposed to function in processes such as membrane transport, signal transduction, and cell adhesion [11,12]. They are enriched in certain lipids (sphingolipids, including gangliosides, sphingomyelin and cholesterol) that show the property of being insol- uble in common detergents, such as Triton X-100. A variety of specific proteins implicated in signal trans- duction from plasma membrane to cytoplasm has also been detected within these domains; some proteins are costitutively enriched within lipid rafts and act through them, and some others are recruited upon specific cell stimulation [12,13]. In this work, we provide evidence that exogenous prosaposin is able to induce ERK phosphorylation and SphK activation, acting through lipid rafts, proba- bly binding different target molecules on the cell plasma membrane. These findings indicate that the proliferative and antiapoptotic functions of prosaposin are dependent on its association with lipid rafts. Results Prosaposin association with lipid rafts in PC12 cells In order to reveal the possible association of prosapo- sin with lipid rafts, we performed immunofluorescence labelling, followed by scanning confocal microscopy analysis. Cells were labelled with antibody against prosaposin (anti-769) and then with cholera toxin B subunit (CTxB), which stains ganglioside GM1 as a specific raft marker [14]. The analysis of prosaposin staining revealed an uneven distribution over the cell surface (Fig. 1), which was also the case for GM1 fluorescence. However, the ganglioside distribution on the cell surface appeared to be highly heterogeneous. The merged image of the two stainings clearly revealed yellow areas, which corre- sponded to colocalization areas between prosaposin and GM1 (Fig. 1A). After triggering with prosaposin, colocalization areas appeared as clusters, indicating raft aggregation in patches upon cell stimulation. As it has been reported that prosaposin triggers a signal cascade after binding to a putative G a0 -coupled cell surface receptor [7], we performed a preliminary analysis of the distribution of G a0 heterotrimeric protein, which might reflect a putative prosaposin receptor, and its association with GM1. In Fig. 1B, the merged image of the staining clearly shows the pres- ence of colocalization areas, indicating, as expected, that G a0 proteins are associated with lipid rafts. After triggering with prosaposin, the labelled molecules, G a0 protein and GM1, do not seem to modify their distri- bution, maintaining an evident colocalization pattern in the merged picture. As prosaposin has been demonstrated to be a ligand for LRP-1 [8], we also decided to evaluate the possible association of this receptor with lipid rafts. Scanning confocal microscopy showed a green fluorescence, cor- responding to LRP-1, unevenly distributed over the cell membrane, similar to the appearance of the red fluorescence corresponding to the raft marker GM1 [14]. The merged image showed two distinct distribu- tion patterns, which revealed the absence of colocaliza- tion areas (Fig. 1C, upper). When the cells were preincubated with prosaposin, a number of yellow- stained colocalization areas became evident (Fig. 1C, lower), indicating recruitment of LRP-1 molecules to lipid rafts. The morphometric analyses aimed at evaluation of the percentages of cells showing prosaposin–GM1, G a0 –GM1 and LRP-1–GM1 colocalization (i.e. dis- playing yellow staining) are reported in Fig. 1D. Preferential association of prosaposin and LRP-1 with lipid raft fractions To analyse the presence of prosaposin and the related receptor in lipid raft fractions of PC12 cells, we investi- gated the distribution of these proteins in fractions obtained by a 5–40% linear sucrose gradient, in either the absence or the presence of triggering with prosa- posin for 5 min (Fig. 2) or 10 min (data not shown). Western blot results revealed that, in untreated cells, prosaposin was enriched mainly in fractions 5 and 6. After treatment, an increase of prosaposin content in these fractions was observed (Fig. 2A). A distribution in the same fractions was observed for G a0 protein in untreated as well as in prosaposin-treated cells (Fig. 2B). When PC12 cell fractions were analysed for LRP-1 distribution (Fig. 2C), a band of 85 kDa, rec- ognized by a monoclonal antibody against LRP-1, was present in Triton X-100-soluble fractions (10 and 11); but, when cells were incubated with prosaposin, LRP-1 clearly switched to Triton X-100-insoluble fractions (4–6). As a control, we also analysed the distribution of calnexin in both treated and untreated cells. As expected, it was found in the fractions corresponding to heavy membranes (Fig. 2D). In contrast, GM1 was Prosaposin signalling through lipid rafts M. Sorice et al. 4904 FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS present in the raft fractions, independently of pros- aposin treatment, as revealed by TLC immunostaining, using a monoclonal antibody against GM1 (Fig. 2E). Prosaposin induces ERK and SphK activation through lipid rafts We previously demonstrated that prosaposin induced ERK phosphorylation and SphK activation in PC12 cells [5]. To investigate the contribution of lipid rafts to the prosaposin effect, we analysed ERK and SphK activation following prosaposin stimulation, in either the presence or the absence of pretreatment with methyl-b-cyclodextrin (MbCD) or filipin III, as these treatments induce cholesterol efflux from the plasma membrane and, consequently, lipid raft disruption [15], or the ceramide analogue d-threo-1-phenyl-2-decanoyl- amin-3-morpholino-propanol (D-PDMP), which inhib- its glucosylceramide synthase and thus leads to extensive depletion of endogenous glycosphingolipids [16]. In Fig. 3A, a western blot analysis shows that pretreatment with MbCD or filipin III, as well as D-PDMP, partially prevented the prosaposin-induced ERK activation. In parallel experiments, ERK phos- phorylation by prosaposin was partially prevented by previous incubation of the cells with a natural ligand of LRP-1, LDL, as well as with pertussis toxin (PT), which catalyses ADP-ribosylation of several G-proteins [17] (Fig. 3B), indicating that LRP-1 receptor, as well as the G a0 -coupled prosaposin receptor, are involved in ERK phosphorylation. As a loading control for immunoblotting experiments, an antibody against b-actin was used. Moreover, basal SphK activity was improved by prosaposin, and this effect was strongly inhibited (about 65%) by pretreatment with MbCD (Fig. 4A). AB CD 80 MergeGM1LRP-1 Merge GM1 Prosaposin +prosaposin Control+prosaposin Control +prosaposin Control Merge GM1 Gα0 Prosaposin/GM1 Gα0/GM1 LRP-1/GM1 Control +prosaposin Control +prosaposin Control +prosaposin 70 60 50 Cells with yellow staining (%) 40 30 20 10 0 Fig. 1. Scanning confocal microscopic analysis of prosaposin, G a0 or LRP-1 association with the raft marker GM1 on the PC12 cell surface. Cells were analysed in the absence or in the presence of prosaposin incubation (10 n M for 5 min at 37 °C), and then labelled with antibody against prosaposin (anti-769), antibody against G a0 or antibody against LRP-1, followed by FITC-conjugated anti-rabbit or anti-mouse IgG. Then, cells were stained with Texas-red-conjugated CTxB (as GM1 staining). One representative cell is shown. (A) Prosaposin–GM1 associa- tion in untreated (control) and treated (+ prosaposin) cells. (B) G a0 –GM1 association in untreated (control) and treated (+ prosaposin) cells. (C) LRP-1–GM1 association in untreated (control) and treated (+ prosaposin) cells. (D) Morphometric analysis of data obtained by confocal microscopy. Two hundred cells were counted for each sample. Histograms indicate the percentage of cells with prosaposin–GM1, G a0 –GM1 and LRP-1–GM1 colocalization. M. Sorice et al. Prosaposin signalling through lipid rafts FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS 4905 In order to confirm these findings, using a highly specific antibody against SphK1, untreated and prosa- posin-stimulated cells were immunoprecipitated with an antibody against SphK1 and analysed by western blot with an antibody against phosphoserine (p-Ser) (Fig. 4B). This revealed that pretreatment with MbCD or filipin III almost completely abolished the prosapo- sin-induced SphK1 activation, indicating that the pro- saposin-triggered signalling pathway occurs through lipid rafts. Moreover, the results also suggested that the ERK pathway may act upstream of SphK activa- tion, as the MEK inhibitor PD98059 prevented pro- saposin-induced SphK1 activation. Protective effect of prosaposin against cell apoptosis occurs through lipid rafts To verify whether the protective effect of prosaposin against cell apoptosis involved lipid rafts, cells were incubated with staurosporine, in either the presence or the absence of prosaposin, with or without pretreat- ment with 5 mm MbCD, and stained with Hoe- chst 33258 (Fig. 5A). The nuclei of control (a) as well as prosaposin-stimulated (c) PC12 cells were stained uniformly with this dye, whereas treatment with MbCD alone showed only a negligible effect on cell apoptosis (e). As expected, treatment of cells with staurosporine caused nuclear condensation and fragmentation (b), whereas in cells treated with prosaposin plus stauro- sporine, a decrease in the number of apoptotic cells was observed (d). When cells were pretreated with MbCD and incubated with prosaposin and stauro- sporine (f), the protective effect of prosaposin against apoptosis was partially inhibited, indicating that the signalling pathways involved in cell death prevention triggered by prosaposin occur through lipid rafts. In addition, DNA fragmentation was quantified as a hypodiploid peak by cytofluorimetric analysis (Fig. 5B). As reported previously [5], a much larger hypodiploid peak was observed after 24 h of stauro- sporine incubation (37.4 ± 4%), as compared to untreated cells (4.9 ± 0.8%); preincubation of cells with prosaposin provided protection against stauro- sporine-induced apoptosis, as indicated by the decrease in hypodiploid cell number (14.8 ± 2%). This effect was significantly inhibited (P < 0.001) when cells were pretreated with MbCD and then incubated with prosaposin and staurosporine (35.3 ± 2.8%). Moreover, we evaluated the early stages of programmed cell death by analysing fluorescein isothio- cyanate (FITC)-conjugated annexin V ⁄ propidium iodide (PI) staining of treated and untreated cells by flow cytometry (Fig. 5C). Once again, it was confirmed that preincubation with MbCD partially prevented the protective effect of prosaposin against apoptosis. It is of note that the analysis of MbCD-treated cells revealed annexin V binding to less than 20% of cells. Thus, pretreatment with MbCD, under our experi- mental conditions, has to be considered as a ‘mild’ A B C D E Fig. 2. Subcellular distribution of prosaposin, G a0 and LRP-1 in PC12 sucrose gradient membrane fractions. Cells, untreated or treated with 10 n M prosaposin, were lysed in lysis buffer, and the supernatant (postnuclear fraction) was subjected to sucrose density gradient separation. After centrifugation, the gradient was fraction- ated, and each fraction was analysed by western blotting with anti- body against prosaposin (A), G a0 (B), LRP-1 (C) or calnexin (D). Alternatively, gangliosides were extracted from each fraction and separated by HPTLC, using silica gel 60 HPTLC plates and chloro- form ⁄ methanol ⁄ 0.25% aqueous KCl (5 : 4 : 1 v ⁄ v ⁄ v) as eluent sys- tem. The plates were immunostained with monoclonal antibody against GM1 (GMB16) (E). Prosaposin signalling through lipid rafts M. Sorice et al. 4906 FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS treatment, without severe changes in membrane orga- nization or signs of distress in the cells, according to Ottico et al. [18]. Discussion In this work, we provide evidence that exogenous pro- saposin binds its target molecule on lipid rafts of the cell plasma membrane and induces ERK phosphoryla- tion and SphK activation through these microdomains. These findings indicate that the trophic and antiapop- totic signalling pathways triggered by prosaposin occur through lipid rafts. The presence of prosaposin within lipid rafts is sup- ported by two independent experimental approaches, i.e. scanning confocal microscopy and sucrose gradient in the presence of Triton X-100, taking advantage of the insolubility of these microdomains in the detergent. Our findings revealed significant colocalization areas between prosaposin and ganglioside GM1, which, although it represents a minor ganglioside component in these cells [19], is a well-known marker of lipid rafts [14]. This prosaposin distribution pattern was confirmed by western blot analysis of sucrose gradient Triton X-100 fractions, which revealed that prosaposin was enriched in Triton X-100-insoluble fractions. After treatment, an increase in prosaposin content in raft fractions was observed. These experiments indicate that prosaposin is associated with lipid rafts, where it may bind different molecules, including a G a0 -coupled receptor and LRP-1. As already reported [20], hetero- trimeric G-proteins have also been detected in lipid rafts; thus, the analysis of G a0 -protein distribution on PC12 cells, as expected, clearly showed the presence of colocalization areas with raft marker molecules, indi- cating stable G a0 -protein localization within lipid rafts. The same result was confirmed by western blot analy- sis of sucrose gradient Triton X-100 fractions, consis- tent with the view that the putative prosaposin receptor is G a0 -coupled [7]. Interestingly, our findings show recruitment of LRP-1 molecules to lipid rafts only after prosaposin triggering. This finding is in agreement with the observation that LRP-1 associates transiently with lipid rafts and that its distribution into membrane microdomains is cell type specific [21]. These multifunctional lipoprotein receptors are estab- lished cargo transporters, but their expression at the cell surface and agonistic binding of diverse biological ligands are now thought to potentially evoke signalling pathways involved in cell fate determination [22]. In addition, our data suggest that the LRP-1 receptor, as well as the G a0 -coupled receptor, may play a role in the prosaposin-triggered signal pathway leading to ERK phosphorylation. Indeed, in other cell systems, activation of the MEKK–JNK–cJun signalling cascade and of the Mek1–ERK1 ⁄ 2 pathway by LRP-1 have been reported [23]. However, the main finding of this study is the demonstration that prosaposin induces ERK phos- phorylation and SphK activation through lipid rafts. In previous studies, we demonstrated that prosaposin treatment induces PC12 entry in the S-phase of the cell cycle and prevents apoptosis by activation of ERKs and SphK in PC12 cells [5], as well as in A B Fig. 3. Effect of MbCD, filipin III and D-PDMP on ERK phosphorylation induced by prosaposin. (A) PC12 cells, untreated or treated with 10 n M prosaposin, were lysed in lysis buffer and analysed by western blot with monoclonal antibody against phospho- p44 ⁄ p42. A representative example of three experiments is shown. (B) PC12 cells, prein- cubated with LDL (natural ligand of LRP-1) (10 lgÆmL )1 per 10 6 cells for 30 min at 4 °C) or with PT (inhibitor of G a0 -coupled receptor function) (100 ngÆmL )1 for 30 min at 37 °C), were treated with 10 n M prosa- posin, as above, lysed in lysis buffer, and analysed by western blot with monoclonal antibody against phospho-p44 ⁄ p42. A repre- sentative example of three experiments is shown. Approximately equal protein loading of the gel was verified using an antibody against b-actin. M. Sorice et al. Prosaposin signalling through lipid rafts FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS 4907 different cell types [6]. Here, we confirm and extend these findings, demonstrating the involvement of lipid rafts. Indeed, pretreatment with MbCD or with fili- pin III, which are able to induce cholesterol efflux from the membrane and, consequently, raft disrup- tion, almost completely prevented ERK and SphK activation, with a prosaposin antiapoptotic effect. At the same time, according to Ottico et al. [18], MbCD pretreatment under our experimental conditions does not induce severe changes in membrane organization or signs of distress in the cells, although the lipid domains of plasma membranes lost the ability to sort specific signalling proteins. The finding that pretreat- ment of cells with D-PDMP also inhibited prosapo- sin-induced ERK phosphorylation suggests two different considerations: (a) this finding may consti- tute functional confirmation of the well-known ability of prosaposin to bind gangliosides [9]; and (b) as gangliosides are well-known components of lipid rafts [14], this result strongly supports the view that pro- saposin-triggered signal transduction occurs through lipid rafts. Moreover, we provide evidence that pro- saposin also induces SphK activation through lipid rafts. Most cells express both SphK1 and SphK2; our findings suggest that in our system prosaposin may act through SphK1, as detected by a highly spe- cific antibody. The involvement of SphK-1 activation in our system is in agreement with the notion that the activation of SphK-1 by external stimuli, includ- ing growth factors, results in accumulation of intra- cellular S1P, and, consequently, increased cell proliferation and suppression of apoptosis [24]. Interestingly, our findings suggest that the ERK pathway is upstream of SphK activation, as the MEK A B Fig. 4. Effect of MbCD on prosaposin- induced SphK activation. (A) Cells were preincubated with 5 m M MbCD and stimu- lated with 10 n M prosaposin. After cell lysis, cytosolic fractions were prepared and SphK activity was measured in the supernatant by incubating 50 l M sphingosine-b-octylgluco- side and [ 32 P]ATP[cP]. The modulation of SphK activity by MbCD is evaluated as pmo- les of N-caproyl-S1P. Error bars represent standard deviation. (B) Effect of MbCD and filipin III on SphK phosphorylation induced by prosaposin. Cells, untreated or treated with 10 n M prosaposin, were lysed in lysis buffer and immunoprecipitated with a rabbit polyclonal antibody against SphK1 (M-209) or, as a control, with rabbit IgG with irrele- vant specificity. The immunoprecipitates were analysed by western blot with mono- clonal antibody against p-Ser. The antibody against p-Ser was stripped, and the mem- brane was then reprobed with goat poly- clonal antibody against SphK1 (M-13). A representative example of three experi- ments is shown. Prosaposin signalling through lipid rafts M. Sorice et al. 4908 FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS inhibitor PD98059 was able to inhibit the prosaposin- induced SphK1 activation. However, we cannot exclude the possibility that additional transduction pathway(s) triggered by prosaposin may be responsible for SphK activation. Taken together, these findings point to a role of lipid rafts in the prosaposin-triggered signalling path- way, thus supporting a role for this factor as a new component of the multimolecular signalling complex involved in the neurotrophic response. Further studies are in progress in order to purify and sequence the putative prosaposin receptor(s). Experimental procedures Materials and cells Milk prosaposin was prepared as previously reported [25]. An antibody against the active 22-mer prosaposin peptide (anti-769) [1] and a monoclonal antibody against prosapo- a b c d e A Control +prosaposin +MβCD +MβCD+prosaposin+STS +prosaposin + STS +STS B C f Fig. 5. Effect of Mb CD on the antiapoptotic activity of prosaposin. (A) PC12 cells, incubated with or without 5 mM MbCD, in the presence or absence of 10 n M prosaposin, were treated with 1 lM staurosporine (STS) and stained with Hoechst 33258. Nuclei of control cells were stained uniformly with Hoechst (a), as well as those of cells treated with prosaposin alone (c). Treatment of cells with staurosporine caused nuclear fragmentation and condensation (b). Pretreatment of cells with prosaposin prevented apoptosis (d). This effect was partially inhibited by preincubation with MbCD (f). Treatment with MbCD alone did not show any effect on cell apoptosis (e). A representative example of three independent experiments. (B) Flow cytometry analysis of hypodiploid cells. Histograms represent the percentage of hypodiploid peaks, as detected by PI staining. Mean ± standard deviation of five independent experiments. STS versus prosaposin + STS, P < 0.001; MbCD + prosaposin + STS versus prosaposin + STS, P < 0.001. (C) Flow cytometry analysis of early stages of cell apoptosis by annexin V ⁄ PI staining. Histograms represent the percentage of annexin V-positive cells. Mean ± standard deviation of five independent experiments. STS versus prosaposin + STS, P < 0.001; MbCD + prosaposin + STS versus prosaposin + STS, P < 0.001. M. Sorice et al. Prosaposin signalling through lipid rafts FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS 4909 sin [26] were employed. A rabbit polyclonal antibody against SphK1 (M-209) and a goat polyclonal antibody against SphK1 (M-13) (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) were employed for immunoprecipi- tation and detection of SphK1 in western blots. The rat pheochromocytoma cell line PC12 was cultured as previously described [5]. Experiments investigating the signalling pathway triggered by prosaposin were performed with or without pretreatment with LDL (Sigma Chemical Co., St Louis, MO, USA, or purified from human plasma as described in [6]), as a natural ligand of LRP-1, or PT (recombinant holotoxin; Calbiochem, San Diego, CA, USA), which binds the G a0 -coupled molecules. In experi- ments aimed at investigating the signalling pathway trig- gered by prosaposin through lipid rafts, cells were pretreated with MbCD (5 mm for 30 min at 37 °C), fili- pin III (10 lm for 30 min at 37 °C) or D-PDMP (30 lm for 5 days at 37 °C) (all purchased from Sigma). Optimal con- centrations and incubation times were carefully checked on the basis of morphological and viability tests. In particular, a Trypan blue exclusion test revealed incorporation into less than 20% of cells pretreated with MbCD or filipin III. The effect of D-PDMP on glycosphingolipid depletion and cholesterol distribution was checked as previously described [27] (data not shown). Immunofluorescence staining PC12 cells were cultured onto coverslips and maintained in DMEM, containing 2% horse serum, for 24 h before pro- saposin treatment (10 nm for 5 min at 37 °C) and labelling. Cells were rinsed, and fixed in 4% formaldehyde in NaCl ⁄ P i for 30 min at 4 °C. After washes, cells were incu- bated for 1 h with rabbit polyclonal antibody against G a0 (anti-769) (Santa Cruz Biotechnology) or mouse mono- clonal antibody against LRP ⁄ a2MR (Immunological Sciences), and this was followed by addition for 45 min of Texas red-conjugated antibody against rabbit or mouse IgG (Calbiochem), and addition of FITC-conjugated CTxB (Molecular Probes, Invitrogen Corps, Carlsbad, CA, USA). Cells were analysed, and images were collected as previ- ously described [28]. Morphometric analyses in double labelling experiments were carried out by evaluating at least 200 cells for each sample, at the same magnification (·630). Isolation and analysis of lipid raft fractions Lipid raft fractions from PC12 cells, untreated or treated with prosaposin (10 nm for 5 or 10 min at 37 °C), were iso- lated as previously described [29], and the fractions were sub- jected to western blot analysis. Samples were normalized for cell number. Blots were probed with monoclonal anti- body against prosaposin, monoclonal antibody against LRP ⁄ a2MR, antibody against G a0 , or polyclonal antibody against calnexin (Sigma) overnight at 4 °C. All the antibody dilutions for blotting and intermediate washes were performed with NaCl ⁄ Tris, containing 0.05% Tween-20. Antibody binding was detected with ECL Tm peroxidase antibody against rabbit or mouse IgG (Amersham Bio- sciences, Amersham, UK), and immunoreactivity assessed by chemiluminescence. Alternatively, gangliosides were extracted from each fraction according to Svennerholm & Fredman [30] and separated by HPTLC, using silica gel 60 HPTLC plates (Merck, Darmstadt, Germany). Chromato- graphy was performed in chloroform ⁄ methanol ⁄ 0.25% aqueous KCl (5 : 4 : 1 v ⁄ v ⁄ v). The plates were immuno- stained for 1 h at room temperature with GMB16 monoclo- nal antibody against GM1 (Seikagaku Corp, Chuo-ku, Tokyo, Japan) and then with horseradish peroxidase (HRP)- conjugated anti-mouse IgM (Sigma). Immunoreactivity was assessed by chemiluminescence. Analysis of ERK and SphK activation PC12 cells (6.0 · 10 7 ), washed in serum-free medium, were incubated with prosaposin (10 nm for 5 min at 37 °C) or, as a positive control, with 4a-phorbol 12-myristate 13-ace- tate (Sigma) (50 ngÆmL )1 for 2 min at 37 °C) (not shown), in DMEM medium containing 2% horse serum. In parallel experiments, cells were preincubated with MbCD (Sigma) (5 mm for 30 min at 37 °C), with filipin III (Sigma) (10 lm for 30 min at 37 °C), with D-PDMP (Sigma) (30 lm for 5 days at 37 °C), with LDL (10 lgÆmL )1 per 10 6 cells for 30 min at 4 °C) or with PT (100 ngÆmL )1 , for 30 min at 37 °C) in the same medium, in either the presence or the absence of prosaposin. Cells were washed twice with ice- cold NaCl ⁄ P i and analysed for ERK and SphK, as previ- ously described [6]. A loading control was performed using a monoclonal antibody against b-actin (mouse IgG 1 , clone AC15, Sigma) and HRP-conjugated anti-mouse IgG (Sigma). Alternatively, cell-free lysates from unstimulated PC12 or PC12 cells stimulated with prosaposin, in the presence or absence of 50 lm MEK inhibitor PD98059, filipin III or MbCD, were immunoprecipitated with a rab- bit polyclonal antibody against SphK1 (M-209) (Santa Cruz). In brief, cells were lysed in lysis buffer, including Na 3 VO 4 , phenylmethanesulfonyl fluoride and protease inhibitors. To preclear nonspecific binding, cell-free lysates were mixed with protein G–Sepharose beads (Bio-Rad, Hercules, CA, USA) and stirred in a rotary shaker for 1 h at 4 °C. After centrifugation (1000 g for 6 min), the super- natant was immunoprecipitated with antibody against SphK1. The immunoprecipitates were subjected to 10% SDS ⁄ PAGE. The proteins were electrophoretically trans- ferred to a nitrocellulose membrane (Bio-Rad), and then, after blocking with NaCl ⁄ Tris containing Tween-20 and 5% milk, probed with monoclonal antibody against p-Ser (Sigma Aldrich) or with antibody against SphK1, respectively. Bound antibodies were visualized with Prosaposin signalling through lipid rafts M. Sorice et al. 4910 FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS HRP-conjugated anti-mouse IgG (Sigma Aldrich), and immunoreactivity was assessed by the chemiluminescence reaction using the ECL western blotting system (Amersham). To confirm the positive band as SphK1, the antibody against p-Ser was stripped from the nitrocellulose, and the membrane was then reprobed with goat polyclonal antibody against SphK1 (M-13) (Santa Cruz) as a loading and reactivity control. Evaluation of cell death Subconfluent PC12 cells, incubated in either the presence or the absence of 10 nm prosaposin for 30 min, were treated with 1 lm staurosporine (Sigma) for 24 h at 37 °C. In par- allel experiments, cells were preincubated with 5 mm MbCD for 30 min at 37 °C. Apoptosis was measured by both morphological analysis and flow cytometry. Morphological analysis of the nuclei was performed by staining the cells with Hoechst 33258 (Sigma), 5 lgÆmL )1 , in 30% glycerol ⁄ NaCl ⁄ P i for 20 min. Cells were examined in an inverted fluorescence microscope (320 nm UV excitation). Viable cells were identified by their intact nuclei, and fragmented or condensed nuclei were scored as apoptotic. DNA fragmentation consistent with apoptosis was quantified as a hypodiploid peak by PI stain- ing and cytofluorimetric analysis, as previously described [31]. Alternatively, apoptosis was also quantified by flow cytometry after double staining using FITC-conjugated annexin V ⁄ PI apoptosis detection kit (Eppendorf, Milan, Italy), which allows discrimination between early apoptotic, late apoptotic and necrotic cells. In this case, in order to detect the early stages of apoptosis, cells incubated in either the presence or the absence of 10 nm prosaposin for 30 min were treated with 1 lm staurosporine (Sigma) for 3 h at 37 °C. Again, in parallel experiments, cells were preincubated with 5 mm MbCD for 30 min at 37 °C. Acknowledgements This work was supported by a grant to RM from Uni- versity ‘Sapienza’ School of Medicine, Rome, Italy. References 1 O’Brien JS, Carson GS, Seo HC, Hiraiwa M & Kishimoto Y (1994) Identification of prosaposin as a neurotrophic factor. Proc Natl Acad Sci USA 91, 9593– 9596. 2 O’Brien JS, Carson GS, Seo HC, Hiraiwa M, Weiler S, Tomich JM, Barranger JA, Kahn M, Azuma N & Kishimoto Y (1995) Identification of the neurotrophic factor sequence of prosaposin. FASEB J 9, 681–685. 3 Hineno T, Sano A, Kondoh K, Ueno S, Kakimoto Y & Yoshida K (1991) Secretion of sphingolipid hydrolase activator precursor, prosaposin. Biochem Biophys Res Commun 176, 668–674. 4 Hiraiwa M, Taylor EM, Campana WM, Darin SJ & O’Brien JS (1997) Cell death prevention, mitogen-acti- vated protein kinase stimulation, and increased sulfatide concentrations in Schwann cells and oligodendrocytes by prosaposin and prosaptides. Proc Natl Acad Sci USA 94, 4778–4781. 5 Misasi R, Sorice M, Di Marzio L, Campana WM, Mol- inari S, Cifone MG, Pavan A, Pontieri GM & O’Brien JS (2001) Prosaposin treatment induces PC12 entry in the S phase of the cell cycle and prevents apoptosis: activation of ERKs and sphingosine kinase. FASEB J 15, 467–474. 6 Misasi R, Garofalo T, Di Marzio L, Mattei V, Gizzi C, Hiraiwa M, Pavan A, Cifone MG & Sorice M (2004) Prosaposin: a new player in the cell death prevention of U937 monocytic cells. Exp Cell Res 298, 38–47. 7 Hiraiwa M, Campana WM, Martin BM & O’Brien JS (1997) Prosaposin receptor: evidence for a G-protein- associated receptor. Biochem Biophys Res Commun 240, 415–418. 8 Hiesberger T, Huttler S, Rohlmann A, Schneider W, Sandhoff K and Herz J (1998) Cellular uptake of sapo- sin (SAP) precursor and lysosomal delivery by the low density lipoprotein receptor-related protein (LRP). EMBO J 17, 4617–4625. 9 Hiraiwa M, Soeda S, Kishimoto Y & O’Brien JS (1992) Binding and transport of ganglioside by prosaposin. Proc Natl Acad Sci USA 89, 11254–11258. 10 Misasi R, Hiraiwa M, Sorice M, Garofalo T, Pavan A, Pontieri GM & O’Brien JS (2000) Functional role of ganglioside–prosaposin interaction in neuronal cells. Recent Res Dev Neurochem 3, 301–316. 11 Ledesma MD, Simons K & Dotti CG (1998) Neuronal polarity: essential role of protein–lipid complexes is axo- nal sorting. Proc Natl Acad Sci USA 95, 3966–3971. 12 Simons K & Toomre D (2000) Lipid rafts and signal transduction. Nat Rev Mol Cell Biol 1, 31–41. 13 Parolini I, Sargiacomo M, Lisanti MP & Peschle C (1996) Signal transduction and glycophosphatidylinosi- tol-linked proteins (lyn, lck, CD4, CD45, G-proteins, and CD55) selectively localize in Triton-insoluble plasma membrane domains of human leukemic cell lines and normal granulocytes. Blood 87, 3783–3794. 14 Parton RG (1994) Ultrastructural localization of gan- gliosides; GM1 is concentrated in caveolae. J Histochem Cytochem 42, 155–166. 15 Ravichandran R & Divakar S (1998) Inclusion of ring A of cholesterol inside the b-cyclodextrin cavity: evidence from oxidation reactions and structural stud- ies. J Incl Phenom Mol Recognit Chem 30, 253–270. 16 Inokuchi J & Radin NS (1987) Preparation of the active isomer of 1-phenyl-2-decanoylamino-3-morpholino- M. Sorice et al. Prosaposin signalling through lipid rafts FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS 4911 1-propanol, inhibitor of murine glucocerebroside synthetase. J Lipid Res 28, 565–571. 17 Hewlett EL, Sauer KT, Myers GA, Cowell JR & Guer- rant RL (1983) Induction of a novel morphological response in Chinese hamster ovary cells by pertussis toxin. Infect Immun 40, 1198–1203. 18 Ottico E, Prinetti A, Prioni S, Giannotta C, Basso L, Chigorno V & Sonnino S (2003) Dynamics of mem- brane lipid domains in neuronal cells differentiated in culture. J Lipid Res 44, 2142–2151. 19 Ariga T, Macala LJ, Saito M, Margolis RK, Greene LA, Margolis RU & Yu RK (1988) Lipid composition of PC12 pheochromocytoma cells: characterization of globoside as a major neutral glycolipid. Biochemistry 27, 52–58. 20 Chun M, Liyanage UK, Lisanti MP & Lodish HF (1994) Signal transduction of a G-protein-coupled receptor in caveolae: colocalization of endothelin and its receptor with caveolin. Proc Natl Acad Sci USA 91, 11728–11732. 21 Wu L & Gonias SL (2005) The low-density lipoprotein receptor-related protein 1 associates transiently with lipid rafts. J Cell Biochem 96, 1021–1033. 22 Hertz J & Beffert U (2000) Apolipoprotein E receptors: linking brain development and Alzheimer’s disease. Nat Rev Neurosci 1, 51–58. 23 Lutz C, Nimpf J, Jemy M, Boecklinger K, Enzinger C, Utermann G, Baier-Bitterlich G & Baier G (2002) Evi- dence of functional modulation of the MEKK ⁄ JNK ⁄ c UN signaling cascade by the low density lipoprotein receptor-related protein (LRP). J Biol Chem 277, 43143–43151. 24 Hait NC, Bellamy A, Milstien S, Kordula T & Spiegel S (2007) Sphingosine kinase type 2 activation by ERK- mediated phosphorylation. J Biol Chem 282, 12058– 12065. 25 Hiraiwa M, O’Brien JS, Kishimoto Y, Galdzicka M, Fluharty AL, Ginns EI & Martin BM (1993) Isolation, characterization, and proteolysis of human prosaposin, the precursor of saposins (sphingolipid activator pro- teins). Arch Biochem Biophys 304, 110–116. 26 Stastny JJ, Kishimoto Y, O’Brien JS & Beattie CW (1992) Production and characterization of a monoclo- nal antibody to human saposin C. Hybridoma 11, 351–359. 27 Sottocornola E, Misasi R, Mattei V, Ciarlo L, Gradini R, Garofano R, Berra B, Colombo I & Sorice M (2006) Role of gangliosides in the association of ErbB2 with lipid rafts in mammary epithelial HC11 cells. FEBS J 273, 1821–1830. 28 Misasi R, Sorice M, Garofalo T, Campana WM, Giammatteo M, Pavan A, Hiraiwa M, Pontieri GM & O’Brien JS (1998) Colocalization and complex forma- tion between prosaposin and monosialoganglioside GM3 in neural cells. J Neurochem 71, 2313–2321. 29 Garofalo T, Misasi R, Mattei V, Giammarioli AM, Malorni W, Pontieri GM, Pavan A & Sorice M (2003) Association of the death-inducing signaling complex with microdomains after triggering through CD95 ⁄ Fas. J Biol Chem 278, 8309–8315. 30 Svennerholm L & Fredman P (1980) A procedure for the quantitative isolation of brain gangliosides. Biochim Biophys Acta 617, 97–109. 31 Nicoletti I, Migliorati G, Pagliacci MC, Grignani F & Riccardi C (1991) A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry. J Immunol Methods 139, 271–279. Prosaposin signalling through lipid rafts M. Sorice et al. 4912 FEBS Journal 275 (2008) 4903–4912 ª 2008 The Authors Journal compilation ª 2008 FEBS . Neurotrophic signalling pathway triggered by prosaposin in PC12 cells occurs through lipid rafts Maurizio Sorice 1,2 , Sabrina Molinari 1 , Luisa. against apoptosis was partially inhibited, indicating that the signalling pathways involved in cell death prevention triggered by prosaposin occur through