1st Meeting

Perugia, Italy – May 24/25 – 2002

Sphingolipid Metabolism and Cell Regulation

Yusuf A. Hannun
Department of Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston, South Carolina, USA

Studies over the past 2 decades have underscored the importance of specific sphingolipids such as sphingosine, ceramide, and sphingosine phosphate as bioactive molecules involved in diverse cellular responses. As a consequence, many of the enzymes of sphingolipid metabolism have emerged as regulated components in signaling and cell regulatory pathways. The direct targets for these sphingolipids (such as receptors, kinases, and phosphatases, have also emerged as important signal transducers.
More recent studies have introduced another layer of complexity to the understanding of bioactive sphingolipids which derives from the compartmentalized metabolism and action of these lipids. In our studies, we find that the action of ceramide is dependent on the specific compartment of metabolism and generation. One emerging specific pathway connects de novo synthesis of ceramide to activation of the protein phosphatase PP1 and the dephosphorylation of nuclear SR proteins. This serves as a prototype “singling module”. The versatility of sphingolipid metabolism and its interconnection raise the possibility that these individual modules may allow for “integration” of cell responses. This talk will discuss these specific examples of ceramide function and the emerging concepts of sphingolipid metabolism and function.

Can the nuclear sphingomyelin regulate the fate of the cell?

Albi, E., Serenelli, G., Rossi, G., Cataldi, S., Borio, T., Lazzarini, R. Viola Magni, M.P.
Dipartimento di Scienze Biochimiche e Biotecnologie Molecolari, Sezione di Fisiopatologia, Policlinico Monteluce, Perugia

The presence chromatin phospholipid (PLs) fraction rich in sphingomyelin (SM) and of enzymes including neutral-sphingomyelinase (N-SMase, 1), SM-synthase (2) and reverse SM-synthase (3) related with SM metabolism has been shown. Since the activity of these enzymes may modify the amount of diacylglycerol (DAG) or ceramide inside the nuclei, the present research was aimed at investigating the chromatin SM metabolism in cell proliferation, differentiation and/or in apoptotic process. For the proliferation study, the regenerating liver was used as experimental model; the rats were hepatectomised between 8-10 a.m. and killed 6,12, 18, 24 hrs after operation. As control, sham operated animal were used.
For differentiation and apoptotic studies, the erytroleukemic cells treated with DMSO and/or vitamin D3 were used as experimental model. The murine virus-induced Friend cell (Clone 745A) were grown for 5 days in RPMI 1640 medium containing 10% fetal calf serum at 37°C in 5% CO2 humid atmosphere; experimental samples were cultivated as above in the presence of 1.5% DMSO or 10 mMol vitamin D3 or both . Radioactive SM was used as precursors for evaluating N-SMase and reverse SM-synthase activities and radioactive PC for SM-synthase activity. In rat liver regeneration, the activity of SMase and reverse SM-synthase increases at the beginning of S-phase when the SM-synthase is inhibited, so the chromatin ceramide pool is enriched when the cell proliferation start. The differentiated cells after 5 days of culture in the presence of DMSO are 43% and increase to 74% if vitamin D3 is added. The presence of the vitamin D3 alone induces the cells in apoptosis. The SMase and reverse SMase activities, which enriched the chromatin ceramide pool, increase during the apoptotic process whereas the SM-synthase activity, which enriched the chromatin DAG pool, increase during cell differentiation. These results may lead to the hypothesis of the existence of internal equilibrium between different messengers essential for a cell physiological life.
1) Albi,E. and Viola Magni, M.P. Biochim. Biophys. Res. Commun., 236, 29-33, 1997; 2) Albi,E., and Viola Magni,M.P. FEBS Letters 460, 369-372, 1999

Purification and characterization of a sphingomyelin synthase activity from pseudomonas aeruginosa.

Chiara Luberto, Elizabeth Collins, Samer El-Bawab, and Yusuf A. Hannun
Department of Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston, South Carolina, USA

We report the purification and biochemical characterization of an enzymatic activity that synthesizes sphingomyelin (SM) in cultures of Pseudomonas aeruginosa. SM synthase-like activity was found secreted in the medium in cultures of Pseudomonas aeruginosa, strains PA01 and PAK whereas it could not be detected in cultures of Escherichia coli. Such activity showed a neutral pH optimum. From the medium of PAK cultures, SM synthase was concentrated by precipitation with 60% ammonium sulfate and found to bind to hydrophobic matrices such as butyl and octyl sepharose in the presence of high concentrations of salt. After eluting the protein from octyl sepharose columns, fractions containing peak activity were applied onto blue sepharose, mono P, and finally onto mono S columns. Fractions containing SM synthase activity were then resolved on 8 and 12% polyacrylamide-SDS gels and proteins were visualized by silver staining. In the lanes containing peak SM synthase activity three to four silver stained bands correlated with the enzymatic activity. These bands were therefore cut and sent for peptide sequencing. From the mono P column the isoelectric point of the protein was calculated to be approximately 4.9. Moreover, using a superose 12 gel filtration column the molecular weight was estimated to be approximately 60 to 70 kD. Interestingly, a sphingomyelin hydrolysing activity co-purified with SM synthase throughout the purification protocol. Using the fractions with peak activity obtained from the mono S column it was shown that SM synthesis was detected only when phosphatidylcholine or lysophosphatidylcholine and not phosphorylcholine or CDP-choline were used as substrates. Phosphatidylserine, phosphatidylethanolamine, and phosphatidylinositol were not used as substrates in the presence of NBD-C6-ceramide by the purified transferase activity.

Glycospingolipid antigens involved in cd1 mediated T-cell response: synthesis and preliminary bio-evaluation.

Federica Compostella(a) Laura Franchini(a) Luigi Panza(b) and Fiamma Ronchetti(b)
(a) Dipartimento di Chimica e Biochimica Medica, Università di Milano
(b) Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche, Università del Piemonte Orientale, Novara

CD1 proteins are a family of antigen presenting molecules strictly related in structure and evolutionary origin to the major histocompatibility complex (MHC) encoded molecules. However, in contrast to the MHC encoded products, CD1 molecules restrict T-cells response to non-peptide lipid and glycolipid antigens.
In particular, it has been demonstrated that sulfatide, a mixture of galactosylceramides with different fatty acids bound to the amino group of sphingosine and sulfated in position 3, is the antigen presented by a CD1a protein.
We are involved in a project aimed at elucidating the antigen structural features required for the recognition by CD1, and for T-cells activation; in this context we decided to investigate the influence of variations in the sulfatide lipid tails on the recognition of CD1a-restricted T-cells; thus, we planned the synthesis of various sulfatides, differing in the acyl chain length, which are represented in the natural mixture (Fig. 1).

In the present communication we will describe the synthesis and a preliminary bio-evaluation of such a family of sulfatides.
In order to obtain the sulfatide by synthesis, it is necessary to have access to the aglycone part, the ceramide. In particular, we focused our attention to 3-OH protected azidosphingosine as it is well known that it is a better substrate than ceramide in glycosylation reactions. New syntheses of azidosphingosine starting from chiral pool have been realized. The so obtained sphingosine derivative has been glycosylated with a galactosyl donor and different fatty acids have been introduced. Final sulfation at position 3 of the galactose moiety allowed to obtain the desired sulfatides.
The synthetic compounds have been tested for activation of sulfatide-specific T cells and compared to a commercial sample of sulfatide isolated from bovine brain and containing the natural fatty acids mixture on ceramide moiety. The activity of the synthetic and natural sulfatides was studied through the release of TNF-alpha or IL-4 as T cell activation read out. The results showed that the analyzed T cells are differently activated by the synthetic sulfatides with respect to the natural mixture.

Prion protein-ganglioside association in microdomains of lymphocyte plasma membrane.

V. Mattei, M. Sorice, *A. Pavan
Dip. Medicina Sperimentale e Patologia, Università “La Sapienza”, Roma
*Dip. Medicina Sperimentale, Università di L’Aquila

Prionic protein may be cause a series of fatal encephalopathies such as Creuzfeldt-Jakob disease of humans, scrapie of sheep and bovine spongiform encephalopathy. All protein diseases are characterized by the conversion of the constitutive cellular prion protein (PrPC) into the pathogenic form, which is also known as scrapie PrP (PrPSc). Recent evidences showed that PrPC has been detected in human lymphocytes, where the level of PrPC was regulated as a consequence of T cell activation. Since PrpC is bound to the cell surface by glycosylphosphatidylinositol (GPI)-anchored linkage, in this study we evaluated the presence of PrPC in microdomains, which are specialized glycosphingolipid-enriched plasma membrane domains, involved in signal transduction pathway. We preliminarly analyzed protein distribution on the plasma membrane of both lymphocytic and neuronal cells. The results, obtained by scanning confocal microscopy, revealed the presence of large clusters. In order to clarify whether these clusters are corresponding to microdomains, we analyzed PrPC association with specific gangliosides. The presence of PrPC in microdomains was further confirmed by their presence in the Triton-insoluble fractions, obtained by sucrose linear gradient. Additional evidence for a specific PrPC-ganglioside interaction in these cells was derived from the results of TLC analysis, showing that prion protein was associated in PrPC immunoprecipitates with GM3, which is the main ganglioside constituent of microdomains in these cells. These results demonstrate a direct physical association of PrPC with ganglioside GM3 within microdomains of lymphocytic cells, suggesting that PrPC represents a structural component of the multimolecular signaling complex involved in T cell activation and other dynamic lymphocytic plasma membrane functions.

Phospholipase D1 is phosphorylated and activated in human airway epithelial cells stimulated by sphingosine-1-phosphate: role of src tyrosine kinase and protein kinase C-D.

Anna Ghelli, Anna Maria Porcelli, Annalisa Facchini*, Silvana Hrelia*, Flavio Flamigni* and Michela Rugolo
Dipart. di Biologia Ev.Sp. and *Dipart. di Biochimica “G.Moruzzi”, Università di Bologna

Sphingosine-1-phosphate (SPP) belongs to a group of lipid mediators that regulate cell proliferation, differentiation and survival [Hla, T. et al. 1999, Biochem. Pharmacol. 58, 201-207]. SPP exhibits high affinity with the edg subfamily of G protein-coupled receptors [Fukushima, N. et al. 2001, Annu. Rev. Pharmacol. Toxicol. 41, 507-534], which comprises edg-1, edg-3, edg-5 (also called AGR16 or H218), edg-6 and edg-8. These receptors are coupled differentially via Gi, Gq and G12/13 and Rho to multiple effector systems including adenylate cyclase, phospholipase C (PLC) and D (PLD), intracellular Ca2+ elevation, extracellular signal regulated kinases, actin stress fibres remodelling, etc.. In a human airway epithelial cell line (CFNPE9o-), we have previously reported that SPP induced elevation of cytosolic Ca2+ and PLD-dependent activation of stress fibres formation. All these effects were sensitive to pertussis toxin treatment, suggesting the involvement of a Gi type of G protein [Orlati,S. et al. 2000, Arch. Biochem. Biophys. 375, 69-77; Porcelli et al., 2002, Cell Signalling 14, 75-81]. Two mammalian PLD enzymes PLD1 and PLD2 have been cloned and shown to have a broad tissue distribution. Both enzymes prefer phosphatidylcholine as substrate and require phosphatidylinositol 4,5-bisphosphate for activity. PLD1 is activated by low molecular weight G proteins either of the Rho family or ADP ribosylation factor (ARF), by protein kinase C (PKC), and can also be tyrosine phosphorylated. In contrast, purified PLD2 is not stimulated by ARF, Rho and PKCa or b, but it can be tyrosine phosphorylated [Liscovitch, M. et al., 2000 Biochem.J. 345, 401-415]. Despite it is known that PKC is involved in the activation of PLD in intact cells, the role of the different PKC isoforms is still poorly understood.
In this study the regulatory role of the PKC-d isoform was revealed by use of the specific inhibitor rottlerin, in combination with antisense oligodeoxynuclotide to PKCd. Cell treatment with antisense oligodeoxynuclotide, but not with sense oligodeoxynuclotide, completely abolished PKCd expression and caused strong inhibition of SPP-stimulated phosphatidic acid formation. Indeed, among the PKC a, b, d, e and x isoforms expressed in these cells, only PKCd was activated upon cell stimulation with SPP, as indicated by translocation into the Triton X-100-soluble membrane fraction. Furthermore, pertussis toxin and genistein abolished both PKCd translocation and PLD activation. In particular, a significant reduction of phosphatidylbutanol formation by SPP was observed in the presence of PP1, an inhibitor of Src tyrosine kinase, and, furthermore, the activity of Src kinase was increased by SPP and inhibited by PP1. However, the level of PKCd tyrosine phosphorylation was not increased in SPP-stimulated cells, suggesting that Src did not directly phosphorylate PKCd. Finally, the level of serine phosphorylation of PLD1 and PLD2 isoforms was not changed, whereas threonine phosphorylation of the PLD1 isoform only was significantly increased in SPP-treated cells, but not in the presence of rottlerin. In conclusion, in CFNPE9o- cells SPP interacts with an edg-family membrane receptor linked to a Gi type of G protein, leading to activation of the PLD1 isoform, through a signalling pathway involving Src and PKCd.

Inhibition of human PBL proliferation by inhibitors of the sphingolipid pathway.

Lucia Cavallini and Adolfo Alexandre
Dipartimento di Chimica Biologica, Università di Padova

The activation of T lymphocytes is the central event in the development of immunity. IL-2 is produced early upon stimulation, together with the high affinity IL-2 receptor, (CD25). They mediate proliferation and differentiation responses, as well as apoptosis at the end of the process of clonal expansion.
Two independent studies report the involvement of an acidic sphingomyelinase (SMase) in ceramide production in the CD28 costimulary signalling of T cell (Boucher (1995) J Exp Med , 181, 2059), and of a neutral fumonisin B1-sensitive SMase in a T cell hybridoma (Tonnetti (1999) J. Exp. Med. 189, 1581). Since in many instances the destiny of ceramide is to be deacylated to sphingosine and then phosphorylated to sphingosine-1P (SPP) we studied the role of these sphingolipid derivatives on the lymphocyte responses to activating stimuli. We performed experiments to test the effect of FumB1, a sphingosine syntase inhibitor but also neutral sphingomyelinase inhibitor, D-erythro-MAPP an inhibitor of neutral ceramidase and N,N-dimethylsphingosine (DMS), a competitive inhibitor of sphingosine kinase(SPHK) on lymphocyte proliferation induced by phytohemoagglutinin (PHA) and found a strong inhibition by these agents. These findings suggest the involvement of SPP in lymphocyte activation. Inhibition is not relieved by addition of exogenous C2-ceramide, sphingosine or SPP in a large concentration range. All reagents induce a paradoxical increase of inhibition. Sphingosine or ceramide however could act as pro-apoptotic agents and SPP could not enter the cell or act through specific inhibitory receptors (Edg family).
SPP seems slightly increased in PBL after 24 hours of stimulation. To address the involvement of SPP in T cell activation we tried to inhibit SPHK expression by means of five different antisense oligonucleotides directed toward human SPHK isoform-1. An inhibitory effect of 10% was effectively obtained at 10 uM concentration with two of the 24mers tested but the data are not significant due to the interferences of the oligonucleotide content of thimidine in the 3H thimidine incorporation. Other methods less sensitive and reliable as the MTT or Alamar-blue tests tend to confirm this data. We also measured the activation of the SPHK in the extract of PHA stimulated PBL after various times and found a net increase of the SPHK activity at 24 hours. This information together with the increase of SPP and with the inhibition by DMS prove the involvement of SPP in the mitogenic response of T cell.
The late increase of SPP and SPHK activity in PHA stimulated PBL indicate their involvement also in cell cycle control following TCR stimulation. Indeed adding DMS 1 hour after stimulation does not abolish the inhibition of proliferation. Furthermore DMS has no appreciable effect on NFkB activation, and also the production of IL-2 and release of soluble CD25 are only marginally inhibited by DMS (decreases ranging from 20-40% compared to the complete inhibition of the proliferation).
The DMS inhibition (75-100%) of proliferation is also observed in IL-2-dependent proliferation of PBL pre-stimulated with PHA for 74 hours, then washed and rested in low serum. The inhibition is accompanied by inhibition of pRb phosphorylation and cyclin A increase .

Free and cell-associated ganglioside interaction differently affects the biological activity of fibroblast growth factor-2 (FGF-2).

Marco Rusnati, Chiara Urbinati and Marco Presta
Sezione di Patologia Generale, Dipartimento di Scienze Biomediche e Biotecnologie, Facoltà di Medicina, Università di Brescia

Gangliosides, normally associated to cell membranes, are shed in extracellular environment during tumor growth. Also, gangliosides regulate tumor neovascularization that occurs through the stimulation of endothelial cells by angiogenic growth factors such as fibroblast growth factor-2 (FGF-2). In vitro, FGF-2 induces proliferation of endothelial cells by interacting with high affinity tyrosine kinase receptors (FGFRs) and low affinity heparan-sulfate proteoglycans (HSPGs) of the cell surface. Here we report about the ability of both free and cell-associated gangliosides to bind FGF-2 and to modulate its cell-interaction and biological activity.
Free gangliosides: Size exclusion chromatography demonstrates that native, but not heat-denatured FGF-2 binds to GT1b, GD1b or GM1 ganglioside. Also, gangliosides protect native FGF-2 from trypsin digestion (the order of relative potency being: GT1b>GD1b>GM1=GM2>GM3= galactosyl-ceramide) whereas asialo-GM1 (aGM1), free NeuAc, and N-acetylneuramin-lactose are ineffective. Thus, FGF-2 gangliosides interaction occurs through NeuAc residue(s) but it is modulated by the oligosaccharide chain and the ceramidic portion of the glycosphingolipid. Scatchard plot analysis of the binding data of fluorochrome-labeled GM1 (BodyPy-GM1) to immobilized FGF-2 indicates that FGF-2/GM1 interaction occurs with a Kd equal to 1-6 ?M. By evaluating the capacity of synthetic FGF-2 peptide fragments to prevent the binding of BodyPy-GM1 to immobilized FGF-2, the COOH-terminus of FGF-2 has been identified as the putative ganglioside-binding region. Gangliosides inhibit the binding of FGF-2 to FGFRs but not to HSPGs of endothelial GM 7373 cells (the order of relative potency being GT1b>GD1b>GM1>aGM1). Accordingly, gangliosides inhibit the mitogenic activity exerted by FGF-2 in GM 7373 cells. Also in this case, GT1b is the most effective among the gangliosides tested while aGM1, free NeuAc, N-acetylneuramin-lactose, and galactosyl-ceramide are ineffective.
Cell associated gangliosides: Treatment of GM 7373 cells with ganglioside biosynthesis inhibitors fumonisin B1, PDMP or PPPP impairs their capacity to proliferate when exposed to FGF2. Also, the mitogenic activity of FGF2 is inhibited by the GM1-binding cholera toxin B subunit (CTB). Conversely, overloading of GM 7373 cell membranes with exogenous GM1 causes a 10-fold increase of the mitogenic potency of FGF2. 125I-FGF2 binds to cell membrane GM1 (Kd = 3 nM) in complex ganglioside/heparan sulfate-deficient CHO-K1-pgsA745 cell mutants that were overloaded with exogenous GM1. Moreover, FGF2 competes with FITC-CTB for the binding to cell membrane GM1 in different CHO cell lines independently of their capacity to express heparan sulfate proteoglycans. Conversely, CTB inhibits cell proliferation triggered by FGF2 in CHO cells overexpressing the tyrosine kinase FGF receptor-1 (FGFR1). Finally, GM1-overloading confers to FGFR1-transfected, complex ganglioside-deficient CHO-K1 cell mutants the capacity to proliferate when stimulated by FGF2. This is inhibited by CTB. Cell proliferation triggered by serum or by TPA is instead independent of the cell membrane ganglioside milieu.
Here we demonstrate that FGF-2 binds both free and cell-associated gangliosides. The distinct interaction leads however to opposite effects. Indeed, free gangliosides act as FGF-2 antagonists by sequestering the growth factor in the extracellular environment and preventing cell interaction. On the contrary, cell-associated gangliosides act as co-receptors for FGF-2 and their interaction with the growth factor is required for cell proliferation.

Nitric oxide in glioma cell growth: role of ceramide.

P. Viani, P. Giussani, L. Brioschi, L.Riboni, G. Tettamanti
Department of Medical Chemistry and Biochemistry, University of Milan

Many literature reports indicate ceramide and nitric oxide (NO) as crucial modulators of the cell fate in the nervous system. In fact in cells from the nervous system ceramide and NO are both able to regulate important biological events such as cell growth, differentiation and apoptosis possibly acting as modulators of the related signal transduction pathways. Notwithstanding these functional relationships very little is known on the possible role of ceramide in the NO-mediated signaling in the nervous system. In our laboratories we obtained evidence that NO acts as an antiproliferativee agent in C6 glioma cells. This is associated to an increase of cellular ceramide and can be mimicked by treatments that augment intracellular levels of ceramide. Different metabolic studies indicate the NO-mediated ceramide accumulation is due to its reduced utilization for the biosynthesis of complex sphingolipids. Notwithstanding this evidence the enzymatic activity of both sphingomyelin synthase and glucosylceramide synthase are not modified by NO. On these basis we investigated the possibility that the availability of ceramide for the biosynthesis of complex sphingolipids is impaired by NO. The effect of NO on the metabolic utilization of N-hexanoyl- [3H]sphingosine by C6 glioma cells demonstrate that the biosynthesis of N-hexanoyl-complex sphigolipids which does not require ceramide trafficking from endoplasmic reticulum to Golgi apparatus, is not affected. On the contrary the utilization of [3H]ceramide derived from recycled [3H]sphingosine, produced from N-hexanoyl- [3H]sphingosine catabolism, is strongly inhibited by NO. Different experimental approaches that influence the traffic between endoplasmic reticulum and Golgi apparatus indicate that the transport possibly vesicle-mediated of ceramide between these subcellular compartments is impaired by NO. As a consequence newly synthesized ceramide accumulates and appears to act as a mediator of the antiproliferative effect of NO. In conclusion these results demonstrate that in glioma cells, the modulation of ceramide intracellular traffic can contribute to the regulation of the intracellular levels of this sphingoid mediator and participate to ceramide-mediated signaling pathways.

DNA damage in human fibroblasts exposed to fumonisin B1.

F. Galvano(1), L. La Fauci(1), A. Russo(2), A. Campisi2, A. Vanella(2) and M. Renis(2)
(1) Department of Agro-forestry, Environmental Science and Technology, University of Reggio Calabria; (2) Department of Biochemistry, Medical Chemistry and Molecular Biology, University of Catania

Fumonisins are mycotoxins produced by several Fusarium species (Fusarium verticilloides and Fusarium proliferatum) that infest corn and other cereals. Among several classes of fumonisins (A, B, or C) fumonisin B1 (FB1) is the most prevalent and has higher toxicological significance because it is frequently present at high concentrations on both diseased and apparently healthy corn plants. FB1 ingestion as the purified compound as well as F. moniliforme-contaminated corn is linked to a wide spectrum of severe animal diseases and patho-physiologic effects such as equine neurodegenerative disease named “leukoencephalomalacia”, porcine pulmonary oedema, immunosuppression in poultry and nephrotoxicity in sheep, toxicity in fish, atherosclerosis and cardiac thrombosis in non-human primates, cardiovascular toxicity in swine and liver toxicity and carcinogenicity in rats and mice. In addition, epidemiological studies in some areas of South Africa and China evidenced a correlation between the ingestion of FB1 contaminated food and higher incidence of oesophageal carcinoma in humans. FB1 structurally resembles sphingoid bases and acts as inhibitor of ceramide synthetase, a key enzyme involved in de novo sphingolipid biosynthesis and in the reacylation of free sphingoid bases derived from sphingolipid turnover. This inhibitory effect leads to accumulation of free sphinganine and sphingosine and subsequent induction of cell death. However, the down stream effectors activated by these sphingolipids in the cell death-signalling pathway are little known. The aim of this study was to evaluate, in FB1 exposed human fibroblasts, the involvement of oxygen free radicals and of some other biochemical pathways, caspase-3 activity, poly(ADP-ribose)polymerase (PARP) cleavage and DNA damage evaluated by COMET assay. Our results indicate that FB1 treatment (48, 72 h and 10, 50, 100 mM) does not affect cellular viability. Conversely, after 72 h of treatment, FB1 (50 and 100 mM) induced DNA damage, an enhancement of caspase-3-activity and cleavage of poly(ADP-ribose)polymerase compared to controls. In addition, FB1 increased the expression of HSP70 in a concentration and time-dependent manner. Our results indicate that DNA damage of apoptotic type in human fibroblasts is caused by exposure to FB1 at high concentrations and for prolonged time and that the genotoxic potential of FB1 has probably been underestimated and should be reconsidered.

Apoptosis in HT-29 colon cancer cells by ceramide: temporal relationship with impairment of the transport and processing of pro-cathepsin D.

Daniela De Stefanis(1), Patrizia Reffo1, Gabriella Bonelli(1), Francesco M. Baccino(1), Giusy Sala(2), Riccardo Ghidoni(2), Patrice Codogno(3) and Ciro Isidoro(4)
(1) Dipartimento di Medicina e Oncologia Sperimentale, Università di Torino
(2) Laboratorio di Biochimica, Ospedale San Paolo, Università degli Studi di Milano
(3) INSERM U504, Villejuif, France
(4) Università “A. Avogadro”, Dipartimento di Scienze Mediche, Novara

Ceramide has been suggested as an important mediator of apoptosis. The effect of increased intracellular levels of ceramide on the transport and processing of cathepsin D (CD), a lysosomal protease implicated in apoptosis of tumour cells, was investigated in HT-29 human colorectal cancer cells. Lysosomal targeting of proCD was inhibited in a dose- and time-dependent manner by exogenous N-Acetylsphingosine (NAS, also known as C2-ceramide), but not by its C2-dihydroceramide (DH-C2) inactive analogue. This effect of NAS was mimicked, at least in part, by 1-phenyl-2-(decanoylamino)-3-morpholino-1-propanol (PDMP), which inhibits the metabolic conversion of endogenous ceramide into sphingolipids. Both these drugs effectively induced an increase of intracellular ceramide levels of 3.5-folds and of about 8-folds over basal levels within 8 h and 24 h of incubation, respectively. The treatment with either NAS or PDMP revealed to be cytotoxic for HT-29 cells and lead to cell death with classical characteristics of apoptosis including cytosolic release of cytochrome c from mithocondria and cell surface exposition of annexin V binding sites. Morphological signs of apoptosis and DNA fragmentation became apparent, however, only between 24 and 48 h of incubation and poly(ADP ribose)-polymerase cleavage, a hallmark of caspase 3 activity, occurred not earlier than 8 h from incubation. Caspase 3 inhibition protected, whereas CD inhibition by pepstatin A did not, from NAS-induced apoptosis, indicating that once high levels of ceramide have accumulated in the cell CD-mediated proteolysis is not required for subsequent reactions leading to apoptosis. The above-described effects of NAS and PDMP on the transport and maturation of proCD were apparent already 2 h after incubation with the drugs, which is much earlier than when classical biochemical and morphological evidences of apoptosis could be detected. This suggests that impairment of glycoproteins transport along the secretory pathway due to increased levels of cell-associated ceramide is an early event in cells undergoing apoptosis. Of note, secretion of proCD was almost abolished and formation of double-chain mature CD was reduced and delayed by NAS, whereas PDMP did not affect much proCD secretion, though it largely inhibited the endosomal-lysosomal segregation of this molecule.

Serum deprivation increases ceramide levels and induces apoptosis in embryonic hippocampal HN9.10e cells.

Laura Colombaioni(1), Luisella Colombini(2), Laura M. Frago(3), Isabel Varela-Nieto(3), Rossana Pesi(2) and Mercedes Garcia-Gil(2)
(1) Institute of Neurophysiology of CNR, Pisa
(2) Dpt. of Physiology and Biochemistry, University of Pisa
(3) Instituto de Investigaciones Biomedicas Alberto Sols, CSIC-UAM, Madrid 28029, Spain

Sphingolipid metabolites have been involved in the regulation of proliferation, differentiation and apoptosis. While cellular mechanisms of these processes have been extensively analysed in postmitotic neurones, little is known about proliferating neuronal precursors. We have taken as a model of neuroblasts the embryonic hippocampal cell line HN9.10e. In undifferentiated cells, apoptosis was induced by serum deprivation and by treatment
with N-acetylsphingosine (C2-Cer). The intracellular levels of ceramide peaked at one hour following serum deprivation. We observed translocation of Bax from cytosol to mitochondria after one hour of serum withdrawal followed, two hours later, by cytochrome c release from mitochondria. These events occurred without mitochondrial membrane potential loss nor mitochondrial calcium raise. As calcium is implicated in several cell death pathways, we analysed intracellular calcium levels after longer periods of serum deprivation. After six hours, an increase of cytosolic Ca++ was detected in HN9.10e cells loaded with the Ca++ indicator Fluo3-AM. We have also measured caspase-3 activity and we have found caspase-3 activation after 24 hours but not after 4 hours of serum deprivation or C2-Cer treatment. Cells serum-deprived for four hours and then grown in complete medium for twenty hours fully recovered viability. Summarising, in HN9.10e cells, apoptosis involves increases in ceramide levels, translocation of Bax, release of cytochrome c, and maintenance of mitochondrial functionality followed by calcium deregulation. The lack of caspase-3 activation, in addition to the maintenance of mitochondrial function could allow the reversibility of death commitment and provide neuroblasts a longer temporal window to decide their fate.

The proapoptotic effect of retinoic acid in breast cancer cells requires RARb and is mediated by increased endogenous ceramide levels.

Giulia Somenzi(1), Silvia Pozzi(1), Giusy Sala(2), Riccardo Ghidoni(2), Nicoletta Sacchi(1)
Labs of Molecular Genetics(1) and Biochemistry(2),
San Paolo University Hospital, School of Medicine, University of Milan

Retinoids are important regulators of cell proliferation, differentiation and apoptosis. One of the most active retinoids is all-trans-retinoic acid (ATRA) that is being evaluated as chemopreventive and chemotherapeutic agent for many cancers, including breast cancer. It is well known that the molecular action of ATRA is mediated by nuclear receptors (RARs a, b, g and RXRs a, b, g), and that RARb silencing is one of the main reasons for breast cancer resistance towards treatment with retinoids. The ATRA mechanism of action as a proapoptotic agent is largely unknown.
Here we focus on the possibility that endogenous ceramide is one of the molecular targets of ATRA in breast cancer cells, as it was suggested for other cell lines (neuroblastoma and prostate carcinoma cells) and that the mechanism of action is influenced by the nuclear receptor status of the cell line. For this purpose we used two breast cancer cell lines, T47D cells that have no basal expression of RAR?, but that can re-express this receptor after treatment with pharmacological doses of ATRA, and MDA-MB-231 cells, that have a lack of transcription of RAR?, due to high methylation of the CpG islands in its promoter. In addition, T47D cells express high levels of RAR?, that is known to promote RAR? transcription, whereas MDA-MB-231 cells do not.
We found that ATRA (1 ?M, 48 h) induced two-fold endogenous ceramide accumulation in T47D cells, whereas no ceramide increase was observed in MDA-MB-231, after the same treatment. Endogenous ceramide increase correlated with the antiproliferative effect of the drug: ATRA treatment concomitantly impaired cell growth, colony formation and promoted apoptosis. Conversely, no biological effect was observed in MDA-MB-231 cells. Concomitant treatment of ATRA and RARs antagonists resulted in lack of ceramide accumulation and biological effect, thus suggesting that the ATRA signaling pathway requires RAR? expression and is mediated by endogenous ceramide accumulation.

Sialic acid acetylation suppresses the pro-apoptotic activity of GD3.

F. Malisan*, B. Tomassini*, L. Franchi*, N. Ventura*, M.R. Rippo*, I. Condo’*, L. Liberati*, A. Rufini*, C. Nachtigall^, B. Kniep^ and R. Testi*.
*Laboratory of Immunology and Signal Transduction, Department of Experimental Medicine and Biochemical Sciences, University of Rome Tor Vergata, Roma
^ Institute of Immunology, Technical University of Dresden, Germany.

GD3 disialoganglioside acts as an intracellular mediator of apoptosis being responsible for the opening of the mitochondrial permeability transition pore, the release of cytochrome c and the activation of caspase-9. Sialic acid 9-O-acetylation, the most common post-synthetic modification of GD3, affects the physicochemical properties of the parent molecule and is expected to modify its functions. We investigated whether sialic acid 9-O-acetylation affects the pro-apoptotic function of GD3.
Mitochondria are a primary target for GD3. Interestingly, 9-O-acetyl-GD3, unlike GD3, is completely unable to induce cytochrome c release and caspase-9 activation from isolated mitochondria. In addition, whereas exogenous GD3 directly triggers apoptosis of HEK293 cells, exogenous 9-O-acetyl-GD3 can not. Moreover, while most cell types undergo apoptosis upon GD3 synthase overexpression due to endogenous GD3 synthesis and accumulation, the cell line HEK293 is resistant to GD3 synthase-mediated apoptosis. We observed that these cells, upon GD3 synthase overexpression, converted part of the GD3 into 9-O- acetyl-GD3. Finally, when HEK293 cells are co-transfected with GD3 synthase and a 9-O-acetyl-esterase, to prevent 9-O-acetyl-GD3 accumulation, they undergo massive apoptosis.
These data indicate that sialic acid 9-O-acetylation suppresses the pro-apoptotic activity of GD3 and suggest a new role for sialic acid acetylation in the control of the apoptotic program.

Selective induction of cell death by disialogangliosides through opening of the mitochondrial permeability transition pore.

Luca Scorrano(a), Maria Eugenia Soriano(a), Gunter Kirschner(b), Paolo Bernardi (a)
(a) Department of Biomedical Sciences, University of Padova
(b) Fidia Research Laboratories, Abano Terme

Mitochondria amplify cell death signals releasing protein cofactors essential for the activation of effector caspases. Certain lipid second messengers including arachidonic acid and ganglioside GD3 trigger this release by inducing the mitochondrial permeability transition (PT). To gain further insight into the specificity of mitochondrial effects of gangliosides, we compared a and b-series gangliosides, which differ for the number of sialic acid.
The b-series disialogangliosides GD3, GD1b, GT1b and GQ1b induced the PT in isolated mitochondria, while a-series monosialogangliosides GM3, GM2, GM1 and GD1a display no effect. GQ1b and GT1b resulted more potent than GD3 and GD1b. Exogenous GT1b caused PT in situ, accompanied by mitochondrial depolarization and followed by CsA-sensitive cell death. Thus, all b-series gangliosides can act as proapoptotic second messengers that induce apoptosis through the mitochondrial pathway. Our data show that two adjacent sialic acid residues are required for gangliosides to induce the PT. This structural specificity may represent a useful template to tailor drugs aimed at induce cell death.

Effects of ganglioside biosynthesis inhibition on the expression and phosphorylation of membrane tyrosine-kinase receptor p185neu/erbb2 in tumor and normal mammary cell lines.

E. Sottocornola, B. Berra and I. Colombo
Institute of General Physiology and Biological Chemistry, University of Milan

The functional relationship between gangliosides and growth factor receptors (GFRs) with tyrosine-kinase activity has already been extensively analysed. So far, no study has been performed in order to define the relation between gangliosides and the oncoprotein p185neu or its normal counterpart erbB2, although their well known involvement in the mammary adenocarcinoma development and in the mammary gland differentiation.
The mouse mammary adenocarcinoma cell line MG1361, derived from mammary tumors of MMTV-neu transgenic mice, and the normal mouse mammary epithelial cell line HC11, in which erbB2 expression is inducible by the cell confluence degree in culture, are the experimental models analysed in the present study.
In the MG1361 cells, p185neu is constitutively in the phosphorylated active form, because of a single aminoacid substitution in the transmembrane domain (Val664Glu). 10 nM EGF was used to activate (auto-phosphorylation) the erbB2 receptor in HC11 cells. Inhibition of ganglioside biosynthesis in both the cell lines was achieved by 30 mM D-PDMP treatment for five days. Gangliosides were analysed by HP-TLC. p185neu and erbB2 expression and phosphorylation were evaluated by western blot.
Our results demonstrate that, in HC11 cells, erbB2 and ganglioside GM3 expression raise together when increasing the degree of confluence of the cells. In EGF-stimulated HC11 cells, erbB2 is auto-phosphorylated, but its expression level decreases, as well as the ganglioside GM3 content, with respect to not-stimulated cells. Expression of other ganglioside species remains unchanged. Inhibition of ganglioside biosynthesis by D-PDMP treatment in EGF-stimulated HC11 cells does not interfere with the erbB2 auto-phosphorylation ability, but it maintains high expression levels of the receptor. On the contrary, in MG1361 cells, inhibition of ganglioside biosynthesis modifies nor p185neu expression neither its phosphorylation levels.
These data give support to the hypothesis of a tight functional relationship between erbB2 and GM3 in HC11 cells. Both the molecules could co-localise at the plasma membrane level and the presence of GM3 could be of importance in the normal internalisation and turn-over of the activated receptor rather than in the receptor activation. It is to be further investigated if these two molecules are located within the glycosphingolipid enriched microdomains (GEM) and, if it is so, the nature of their molecular relationship.
The absence of a correlation between p185neu and ganglioside expression, in MG1361 cells, might be attributed to the nature of the oncoprotein, that is over-expressed and constitutively activated.

β1,3 galactosyltransferase T5: regulation and involvement in the expression of CA19.9 tumor antigen.

Marco Trinchera
Dipartimento di Scienze Biomediche Sperimentali e Cliniche,Università dell’Insubria, Varese

The CA19.9 antigen is a sialyl-Lewis a tetrasaccharide carried by mucins and glycosphingolipids detected in cancer tissues and cell lines. The b 1,3 galactosyltransferase T5 (b 3Gal-T5), an enzyme preferentially acting on glycosphingolipid substrates in vitro, is studied at the molecular level in this research as a candidate key enzyme in CA19.9 biosynthesis. b 3Gal-T5 transcript was first reported to be well detectable only in adenocarcinoma cell lines expressing CA19.9, but not expressed or present at very low levels in the lines lacking the antigen. More recently, we found that the transcript is strongly down-regulated (over 30-folds) in colon adenocarcinomas versus the normal colon mucosa, and other authors reported that even the enzyme activity is down-regulated in colon adenocarcinomas, but less strongly (about 5-folds). These data are paradoxical since CA19.9 biosynthesis is enhanced in cancer with respect to the normal tissue, and in fact it is widely used as a marker in the diagnosis and follow-up of adenocarcinomas of the gastrointestinal tract. Two possible alternative hypotheses to explain the data are: 1) the presence of a novel b 3Gal-T isoform, unknown at the present, differentially expressed in native cancers, and 2) the possibility that b 3Gal-T5 is not primarily regulated at the transcriptional level. To address this issue, we looked for cellular models resembling native adenocarcinomas and found that the gastric adenocarcinoma cell line MKN-45 expresses low levels of b 3Gal-T5 transcript and no CA19.9 antigen, but strong b 3Gal-T activity. In addition, recombinant MKN-45 clones obtained by Fuc-TIII transfection express very high levels of CA19.9 too. Moreover, in the human pancreas adenocarcinoma cell line BxPc3, CA 19.9 antigen and b 3Gal-T5 activity are both well detected, while b3Gal-T5 transcript is really faint, and in other cell lines where b 3Gal-T5 transcript is more abundant than in BxPc3 cells, the b 3Gal-T5 activity is not higher than in BxPc3. In particular, in recombinant HCT-15 cells expressing high levels of b 3Gal-T5 transcript, due to the strong CMV promoter, b 3Gal-T5 activity is comparable to that measured in BxPc3 cells, and thus much lower than in MKN-45. The above results suggest that transcript levels alone are not predictive of b 3Gal-T activity. To rule out the possibility of multiple b 3Gal-T isoforms, we planned to determine the enzyme kinetics in native or recombinant cell lines as well as in normal colon mucosa or adenocarcinomas. Results available by now show that the Km values calculated with MKN-45 or recombinant cells expressing true b 3Gal-T5 are identical toward three different substrates, as well as toward donor UDP-Gal, and undistinguishable from those reported in human normal mucosa or adenocarcinomas, suggesting that another b3Gal-T isoform is not probably involved. To prove that b 3Gal-T5 alone is responsible for b 3Gal-T activity and CA19.9 biosynthesis in cells where the transcript level is low, we are now trying to suppress it in recombinant MKN-45/Fuc-TIII and BxPc3 clones by expression of an antisense b 3Gal-T5 fragment.

α-galactoside α2,6 sialyltransferase, the enzyme responsible for the α2,6 sialylation of N-acetyllactosaminic chains in glycoproteins and glycolipids, is up-regulated in human cancers.

Fabio Dall’Olio and Mariella Chiricolo
Dipartimento di Patologia Sperimentale, Università di Bologna

α-galactoside α2,6sialyltransferase (ST6Gal I) is the only enzyme which can mediate the transfer of sialic acid in a2,6 linkage to N-acetyllactosamine (Gala1,4GlcNAc), a structure commonly found in glycoproteins and glycosphingolipids [1]. The main regulatory mechanisms of ST6Gal.I expression operate transcriptionally, through the use of distinct promoters which results in the expression of mRNA species differing in the 5′ untranslated regions (reviewed in [2]). In several malignancies, including colon and breast cancer, ST6Gal.I shows a dramatic up-regulation [3]. In colon cancer, where this phenomenon is in part mediated by a transcriptional elevation through the promoter which is usually utilized by liver cells [4,5] , the extent of the upregulation of ST6Gal.I and of the α2,6-linked sialic acid on the cell surface is an indicator of a poor prognosis [6,7]. The relationship between ST6Gal I elevation and neoplastic transformation has been investigated under two aspects. First, the phenotypic changes induced by ST6Gal.I overexpression have been studied upon transfection of the ST6Gal I cDNA in colon cancer cell lines devoid of such enzyme activity [8,9]. The phenotypic changes observed are mild and appear to be dependent on the cell type. Second, the effect of ras and neu oncogenes overexpression on ST6Gal I modulation has been investigated in murine fibroblasts and breast cancer cells respectively. Transfection of both K-ras and H-ras in NIH3T3 fibroblasts results in a dramatic transcriptional activation of the gene and a consequent appearance of α2,6-sialylated glycans at the cell surface. Murine breast cancer cell lines have been obtained from spontaneously developed breast tumors in animals constitutivelly expressing the neu oncogene. Among the cell lines obtained, those expressing neu showed a strong ST6Gal I expression, while revertant cell lines not expressing neu showed negligible ST6Gal I expression. These data indicate that the transcription of ST6Gal I gene is controlled by protein kinase pathways and provide a possible link between ST6Gal I upregulation and invasivity of human cancers.
1. Weinstein J, de Souza e Silva U, Paulson JC, J Biol Chem 257, 13835-44 (1982).
2. Dall’Olio F, Glycoconj J 17, 669-76 (2000).
3. Dall’Olio F, Malagolini N, Di Stefano G, Minni F, Marrano D, Serafini-Cessi F, Int J Cancer 44, 434-9 (1989).
4. Dall’Olio F, Chiricolo M, Lau JT, Int J Cancer 81, 243-7 (1999).
5. Dall’Olio F, Chiricolo M, Ceccarelli C, Minni F, Marrano D, Santini D, Int J Cancer 88, 58-65 (2000).
6. Lise M, Belluco C, Perera SP, Patel R, Thomas P, Ganguly A, Hybridoma 19, 281-6 (2000).
7. Vierbuchen MJ, Fruechtnicht W, Brackrock S, Krause KT, Zienkiewicz TJ, Cancer 76, 727-35 (1995).
8. Dall’Olio F, Chiricolo M, Lollini P, Lau JT, Biochem Biophys Res Commun 211, 554-61 (1995).
9. Dall’Olio F, Mariani E, Tarozzi A, Meneghetti A, Chiricolo M, Lau JT, Facchini A, Glycobiology 7, 507-13 (1997).

Involvement of rat cytosolic sialidase neu2 in differentiation and proliferation of PC12 cells.

A. Fanzani, R. Giuliani, D. Zizioli and S. Marchesini
Sezione di Bicohimica, Dipartimento di Scienze Biomediche e Biotecnologie, Facoltà di Medicina, Università di Brescia

Cytosolic sialidase Neu2 is a protein of 379 amminoacids that is highly expressed in rat skeletal muscle and that contains an enhancer/promoter region which is transcriptionally active in rat L6 myogenic cells.
An increase in cytosolic sialidase transcript has been observed by means of RT-PCR during differentiation of PC12 cells, as well as during maturation of rat cerebellar granule cells, two models of neuronal differentiation. In particular, PC12 cells treated with differentiating agents such as NGF and bFGF, showed an increase of Neu2 mRNA up to 3 and 2 days, respectively. Interestingly, we observed also an increment of the transcript with EGF, a growth factor that induces proliferation. To obtain further informations on the sialidase expression we cloned a fragment of 1.4 kb which is sufficient in rat L6 mioblasts to activate the transcription in transient transfections.
Experiments of transient transfections in PC12 cells have confirmed a trascriptionally activation of the promoter with NGF and bFGF. The sequence analysis of the promoter has shown that the presence of two E-box motifs could be important for the activation, together with others tipical sites involved in NGF activation. The cDNA of Neu2 has been cloned in a inducible vector for the stable transfection of PC12 cells, and we are working to produce a stable cell line for the antisense oligo expression. From these experiments we hope to have informations on the Neu2 involvement in the differentiation process. In order to purify enough Neu2 protein for in vitro enzimatic studies, a fusion protein GST-Neu2 is under preparation.

Ceramide levels are inversely associated with malignant progression of human glial tumors.

Riboni L.(1), Campanella R.(2), Bassi R.(1), Viani P.(1), Tettamanti G.(1)
(1) Dpt. of Medical Chemistry and Biochemistry, University of Milan, LITA-Segrate, Milano
(2)  Neurosurgical Operative Unit, Dpt. of Neurological Sciences, University of Milan, IRCCS Ospedale Maggiore Policlinico, Milano

Astrocytomas represent the most frequent and deadly primary intracranial neoplasias in humans. On the basis of their histopathological and clinical features, these tumors can occur in different stages of malignancy, ranging from low- to high-grade tumors. Although some late stage tumors seem to arise de novo, follow-up studies of patients with low or medium stage disease have shown that most will progress over time to a higher stage, supporting malignant progression in vivo. To determine whether ceramide levels correlate with the malignant progression of human astrocytomas, we investigated these levels in surgical specimens of glial tumors of low- and high-grade malignancy. Tumor samples, obtained from 52 patients who underwent therapeutic removal of primary brain tumors were used. The tumors were classified according to standard morphologic criteria and they were grouped into tumors with low-grade and high-grade of malignancy. Sections of normal brain tissue adjacent to the tumor were also analyzed in 11 of the 52 patients. Ceramide levels were significantly lower in the combined high-grade tumors compared to low-grade ones and in both tumor groups compared to peritumoral tissue. The results obtained indicate an inverse correlation between the amount of ceramide and tumor malignancy evaluated by both histological and biochemical parameters. Moreover, the highest ceramide levels were found in patients with a longer survival time. Altogether, the present findings indicate that ceramide is inversely associated with malignant progression of human astrocytomas and poor prognosis. The down regulation of ceramide levels in human astrocytomas emerges as a novel alteration that may contribute to glial neoplastic transformation. The low ceramide levels in high-grade tumors may underlie their rapid growth and apoptotic resistant features. This study supports the rationale for the potential benefits of a ceramide-based chemotherapy.