Biochimie 83 (2001) 801−809 © 2001 Société française de biochimie et biologie moléculaire / Éditions scientifiques et médicales Elsevier SAS. All rights reserved. S0300908401013177/FLA

Regulation of Paramecium primaurelia glycosylphosphatidyl-inositol biosynthesis via dolichol phosphate mannose synthesis Nahid Azzouza, Peter Gerolda, Mamdouh H. Kedeesa, Hosam Shams-Eldina, Regina Wernera, Yvonne Capdevilleb, Ralph T. Schwarza* a

Zentrum für Hygiene und Medizinische Mikrobiologie, Philipps-Universität Marburg, Robert-Koch-Strasse 17, 35037 Marburg, Germany b Centre de Génétique Moléculaire, UPR 2420, CNRS, Gif-sur-Yvette, France (Received 19 March 2001; accepted 27 March 2001)

Abstract — A set of glycosylinositol-phosphoceramides, belonging to a family of glycosylphosphatidyl-inositols (GPIs) synthesized in a cell-free system prepared from the free-living protozoan Paramecium primaurelia has been described. The final GPI precursor was identified and structurally characterized as: ethanolamine-phosphate-6Manα1-2Manα1-6(mannosylphosphate) Manα14glucosamine-inositol-phospho-ceramide. During our investigations on the biosynthesis of the acid-labile modification, the additional mannosyl phosphate substitution, we observed that the use of the nucleotide triphosphate analogue GTPγS (guanosine 5’-O(thiotriphosphate)) blocks the biosynthesis of the mannosylated GPI glycolipids. We show that GTPγS inhibits the synthesis of dolichol-phosphate-mannose, which is the donor of the mannose residues for GPI biosynthesis. Therefore, we investigated the role of GTP binding regulatory ‘G’ proteins using cholera and pertussis toxins and an intracellular second messenger cAMP analogue, 8-bromo-cAMP. All the data obtained suggest the involvement of classical heterotrimeric G proteins in the regulation of GPI-anchor biosynthesis through dolichol-phosphate-mannose synthesis via the activation of adenylyl cyclase and protein phosphorylation. Furthermore, our data suggest that GTPγS interferes with synthesis of dolichol monophosphate, indicating that the dolichol kinase is regulated by the heterotrimeric G proteins. © 2001 Société française de biochimie et biologie moléculaire / Éditions scientifiques et médicales Elsevier SAS. All rights reserved. dolichol / dolichol monophosphate / Dol-P-man / glycosylphosphatidyl-inositol / Paramecium primaurelia

1. Introduction The structure and biosynthesis of GPI-anchor precursors have been elucidated in protozoa and mammalian cells [1–6]. The evolutionary conserved glycosylphosphatidyl-inositol (GPI) core glycans consist of a trimannosyl glucosamine structure linked to ethanolamine phosphate: ethanolamine-PO4-6Manα1-2Manα1-6Manα1-4GlcNH2. The matured GPI precursor is then transferred to the nascent protein via a pseudo transamidase reaction [7]. Since the synthesis and attachment of GPI is crucial for the maturation and transport of GPI-anchored proteins to the cell surface, it is essential to understand the regulation of the biosynthetic pathway leading to the formation of GPIs. Despite a vast amount of structural data, little is known about the enzymology and regulation of GPI *Correspondence and reprints. E-mail address: [email protected] (R.T. Schwarz). Abbreviations: GPI, glycosylphosphatidyl-inositol; Dol, dolichol; Dol-P, dolichol monophosphate; Dol-P-man, dolicholphosphate-mannose; TLC, thin layer chromatography; TLCK, Nα-tosyl-L-lysine chloromethyl ketone; GTPγS, guanosine 5''O-(thiotriphosphate).

biosynthesis. None of the enzymes involved in the GPI biosynthesis have been purified to homogeneity. However, some biochemical studies have been reported on partially purified GlcNAc-PI deacylase [8] and some genes involved in different steps of GPI biosynthesis have been cloned from mammalian cells, yeast and parasitic protozoa [9–15]. In mammalian cells, GTP but not the nonhydrolysable analogues of GTP can stimulate the deacetylation of GlcNAc-PI in vitro, and the possible role of GTP binding proteins in this stimulation was not ruled out [16, 17]. Some evidence for the involvement of adenylyl cyclase/cAMP system and protein phosphorylation in the regulation of protein glycosylation have been reported. Treatment of JEA-3 choriocarcinoma cells with 8-bromocAMP, a cell permeable analogue of cyclic AMP, increases N-glycosylation by stimulating the dolichol/ dolichol-phosphate pathway [18, 19]. Dol-P-man is the donor for mannose residues not only for GPI biosynthesis but also for N-glycosylation. Banerjee et al. [20] showed that the presence of a cAMP-dependent protein kinases stimulates the activity of Dol-P-man synthesis in vitro. Furthermore, in mammalian cells, stimulation of protein kinase C by phorbol-12-myristate-13-acetate enhances GPI biosynthesis [21].

802 We have described a biosynthetic pathway for a polar glycolipid which is destined to anchor the P. primaurelia GPI-anchored proteins [22, 23]. The core glycan of this glycolipid is modified by mannosyl phosphate which is added to the core glycan in two steps: phosphorylation of the trimannosyl intermediate prior to the addition of the mannose linked to phosphate [23]. We investigated the phosphorylation reaction using different nucleotide triphosphates and their corresponding analogous. We demonstrate that the use of GTPγS interferes with the synthesis of the mannosylated GPI intermediates in vitro. We show that the primary effect of GTPγS is due to the lack of Dol-P-man which is the donor of mannose residues that constitute the core glycan of GPI. Therefore we followed the effect of GTPγS by investigating the synthesis of Dol-P-man and the transfer of mannoses residues from Dol-P-man to GPIs. Two enzymes are involved in Dol-Pman synthesis: CTP dependent dolichol kinase and DolP-man synthase. We show that GTPγS interferes with the synthesis of Dol-P from dolichol. Using cholera and pertussis toxins we prove the involvement of signaltransducing GTP binding proteins Gs and Gi in the regulation of Dol-P-man synthesis. Most of the signal transducing activities of G-proteins are associated with the state of activation of the α subunit, which is involved in GDP/GTP exchange and GTP hydrolysis [24]. Certain activated α subunit isoforms can function in stimulatory (αs) or inhibitory (αi) receptor-coupled pathways for the regulation of adenylyl cyclase [25]. Our data denote that the inhibitory effect of GTPγS is due to the activation of a Gαi subunit acting as a negative regulator for Dol-Pman synthesis.

2. Materials and methods 2.1. Materials Guanosine diphosphate-[3,4-3H]mannose (15.1 Cimmol-1) was from DuPont-New England Nuclear. Porcine Liver Dolichol, [1-3H] (10–20 Ci-mmol-1) and Dolichol, [1-3H]monophosphate (10–20 Ci-mmol-1) were from Biotrend. Vibrio cholerae toxin, Bordetella pertussis toxin, guanosine 5”-O-(3-thiotriphosphate), 8-bromoadenosine 3’:5’-cyclic monophosphate (8-bromo-cAMP) were from sigma. 1-(5-isoquinolinessulfonyl)-2-methylpiperazine, HCl (H-7, dihydrochloride), forskolin 7-deacetyl-6-(N-acetylglycyl)- and staurosporin (Streptomyces sp.) were from Calbiochem. 2.2. Preparation of P. primaurelia lysate Cells of P. primaurelia strain 156 were grown as described [22], supplemented with 10 µg/mL tunicamycin to block the N-glycosylation pathway of proteins and 0.5

Azzouz et al. mM 2-deoxy 2-fluoro-glucose. After 1 h at 37 °C, cells were hypotonically lysed using the method described by Masterson et al. [26]. Briefly, cells were resuspended in ice cold water containing 0.1 mM TLCK and 1 µg/mL leupeptin prior to disruption with 50 strokes in a Dounce homogenizer. An equal volume of 100 mM Na-Hepes (pH 7.4), 50 mM KCl, 10 mM MgCl2, 0.1 mM TLCK, 1 µg/mL leupeptin and 20% glycerol was added to the cell lysate prior to freezing at –80 °C. The cell lysate was used for in vitro labeling experiments. 2.3. Biosynthesis and extraction of GPI glycolipids in vitro The lysate (106 cell eq.) was washed (three times) with 50 mM Na-Hepes (pH 7.4), 50 mM KCl, 10 mM MgCl2, 0.1 mM TLCK and 1 µg/mL leupeptin (buffer A). After centrifugation (Beckman J21, 10 000 rpm, 20 min at 4 °C), the pellet was resuspended in buffer A supplemented with 5 mM MnCl2, 0.2 µg/mL tunicamycin (Calbiochem), 1 mM ATP, 1 mM CoA and 2 µCi of [3H]-labeled nucleotide sugars or 1 µCi [3H]-labeled dolichol or 1 µCi [3H]-labeled dolicholmonophosphate. [3H]-labeled dolichol and [3H]-labeled dolicholmonophosphate were solubilized in 1% Triton X-100. The final concentration of Triton X-100 in incubation mixtures was 0.1% preventing any transfer of mannose from newly synthesized [3H]Dol-P-man to GPIs [23]. Assays were supplemented with 1 mM UDP-N-acetyl glucosamine for experiments involving GDP-[3H]mannose. For experiments involving [1-3H]dolichol, and [1-3H]dolicholmonophosphate, 1 mM GDP-mannose and 1 mM UDPN-acetyl glucosamine were added. After a period of incubation, labeled glycolipids were extracted in one step with chloroform/methanol/water (C/M/W, 10:10:3, by volume) and analyzed by TLC. In experiments involving GTPγS, cholera toxin, pertussis toxin, phosphodiesterase inhibitors and kinase inhibitors, treated and not treated membranes were first incubated with these components for 10 min followed by the addition of radioactivity containing differents labeling ingredients (ATP, CoA, tunicamycin). 2.4. Thin layer chromatography Labeled glycolipids recovered in the C/M/W extracts were dried and partitioned between water and watersaturated n-butanol. The glycolipids recovered in the butanol phases were analyzed on silica 60 plates (Merck), using solvent system A: chloroform/methanol/0.25% KCl (10:10:3, by volume) or B: chloroform/methanol/NH4OH (65:35:5, by volume). After chromatography, the plates were dried and scanned for radioactivity with a Berthold LB 2842 automatic scanner.

Regulation of P. primaurelia glycosylphosphatidyl-inositol biosynthesis

803

3. Results 3.1. Inhibition of Dol-P-man synthesis by GTPγS Glycolipids were labeled in vitro with GDP[3H]mannose using P. primaurelia membranes. Membranes were prepared from cells pre-incubated with tunicamycin and 2-deoxy-2-fluoro-D-glucose. The latter is known to deplete the endogenous pool of Dol-P-man and Dol-P-glucose [27] leading to higher labeling efficiency of Dol-P-[3H]man synthesized via GDP-[3H]mannose. The labeling was stopped by adding C/M (1:1, by volume) to the incubation mixture to give a final concentration of C/M/W (10:10:3, by volume). Thin layer chromatography (TLC) analysis, using silica 60 plates (solvent system A), showed a spectrum of glycolipids (I to VII) and Dol-Pman (figure 1A) labeled with GDP-[3H]mannose. All glycolipids have been structurally identified as glycosylinositol-phosphoceramides [22, 23]. The core glycan of the polar glycolipids I and III were shown to be modified by mannosyl phosphate linked to the mannose adjacent to the non-acetylated glucosamine molecule. Dol-P-man is the donor of all mannose residues that constitute the P. primaurelia GPIs core glycan [23]. In the course of our investigations on the regulation of GPI-anchors biosynthesis, we found that GTPγS interferes with the synthesis of GPIs. Incubation of membranes for 10 min with 4 µM GTPγS, prior to adding GDP-[3H]mannose for additional 30 min, blocks the synthesis of mannose labeled GPIs (figure 1B), as compared to the incubation in the absence of GTPγS (figure 1A). In the same concentration range GTPγS has no effect on the biosynthesis of the early GPI intermediates, GlcNAc-Inositol-P-ceramide and GLcNinositol-P-ceramide glycolipids, (not shown). The incorporation of [3H]mannose into Dol-P-man and GPIs was 80% less efficient in the presence of GTPγS. The effect of GTPγS on mannosylated GPIs was observed at a concentration of about 2 µM, reaching maximum at 5 µM, and could only be partially prevented by adding 10 µM GTP (not shown). GTP was unable to prevent the inhibition by GTPγS probably due to a competitive inhibition of the binding of GDP-man to Dol-P-man synthase as it was shown for P. falciparum [28]. The use of ATPγS in the same concentration range even at higher concentration (10 µM) has no effect on the synthesis of the Dol-P-man [23]. In the presence of GTPγS, no Dol-P-man could be detected in the C/M/W extracts (figure 1B). This led us to conclude that its synthesis is inhibited. As Dol-P-man is the donor for mannose residues that constitute the GPI core glycan, the synthesis of mannosylated GPI intermediates is therefore inhibited. The possibility that GTPγS may interfere with the synthesis of GDP-man could be ruled out, since we supplemented directly with labeled GDP-mannose in these experiments.

Figure 1. GTPγS interferes with synthesis of Dol-P-man. TLC analysis of the P. primaurelia glycolipids: glycolipids were labeled in vitro via GDP-[3H]mannose in the absence (A) or presence of 4 µM GTPγS (B). Glycolipids were then extracted in one step by adding C/M (1:1, by volume) to the incubation mixture to give a final concentration of C/M/W (10:10:3, by volume). The C/M/W extracted glycolipids were dried and partitioned between water and water-saturated n-butanol. The glycolipids were analyzed on silica TLC plates using solvent system A: C/M/0.25% KCl (10:10:3, by volume). Radioactivity was scanned using a TLC-scanner (Berthold LB 2842). The structure of the characterized glycolipids is indicated in A (see [22, 23]). E, ethanolamine phosphate; M, mannose; GalN, N-acetylgalactosamine; P, phosphate; GIPC; glucosamine inositol-phosphoceramide.

3.2. GTPγS interferes with dolichol phosphate synthesis Two enzymes are involved in the synthesis of Dol-Pman: CTP dependent dolichol kinase and Dol-P-man synthase. Therefore, we investigated the effect of GTPγS on these two enzymes. First, we established a TLC system able to separate dolichol, Dol-P and Dol-P-man using

804

Azzouz et al.

Figure 2. Effect of GTPγS on Dol-P synthesis. 1 µCi of Dol or Dol-P was added to P. primaurelia membranes in the presence of 1 mM GDP-mannose and 1 mM UDP-N-acetyl glycosamine. The experiments were done in the absence of GTPγS (A, B) or in the presence of GTPγS (B, D). After a period of incubation of 30 min the C/M/W (10:10:3, by volume) extracts were analyzed on silica TLC plates using solvent system B: C/M/NH4OH (65:35:5, by volume). Radioactivity was scanned using a TLC-scanner (Berthold LB 2842).

solvent system B: CHCl3/CH3OH/NH4OH (65:35:5, by volume). In contrast to Dol-P, which remains closly at the origin (Rf 0.16), the dolichol migrates to the front (Rf 0.7). Both compounds are separated from TLC purified P. primaurelia Dol-P-[3H]man (Rf 0.32). Membranes were then incubated with labeled Dol or Dol-P solubilized in Triton X-100. As shown in figure 2A, in the absence of GTPγS the addition of Dol-P leads to the formation of Dol-P-man but also in the presence of GTPγS (figure 2C), indicating that the Dol-P-man synthase is not affected by GTPγS. Labeling with dolichol in the absence of GTPγS leads to the formation of Dol-P and Dol-P-man (figure 2B). However, in the presence of GTPγS no Dol-P or Dol-P-man could be detected (figure 2D). The data suggest that GTPγS interferes with the synthesis of Dol-P, indicating that dolichol kinase is the enzyme which is affected by GTPγS. Therefore we followed the effect of GTPγS on dolichol kinase by identifying the absence or

the presence of Dol-P-man (Dol-P-man synthesis) after labeling with GDP-[3H]man. Since the labeling with GDP-Man does not require the presence of detergent in contrast to labeling with dolichol. 3.3. The involvement of regulatory ‘Gi’ in the inhibition of Dol-P-man synthesis GTPγS is known to irreversibly activate guanine nucleotide binding proteins [24]. The data directed our investigation towards the possible regulation of the Dol-P-man synthesis by heterotrimeric ‘G’ proteins. We investigated the possible involvement of ‘Gαi’ by adding B. pertussis toxin, known to inactivate the ‘Gαi’ subunit via ADPribosylation [29, 30]. We used in the following experiments membranes which were pretreated with pertussis toxins, activated in the presence of 1 mM dithiothreitol for 10 min, prior to the incubation with 4 µM GTPγS for an

Regulation of P. primaurelia glycosylphosphatidyl-inositol biosynthesis

805

additional 10 min. GDP-[3H]mannose was then added to the incubation mixtures. TLC analysis of glycolipids is shown in figure 3. Pertussis toxin acting as antagonist prevents the inhibition of Dol-P-man synthesis by GTPγS, as determined by the higher incorporation of [3H]mannose into all GPI glycolipids (figure 3C, D), compared to the incubation in the presence of GTPγS (figure 3B). Furthermore, membranes pretreated with the pertussis toxin not only reverse the effect of GTPγS but also stimulate the incorporation of [3H]mannose into GPI glycolipids by a factor of 1.5 at 5 µg/100 µL (figure 3C) and by a factor of 5 at 10 µg/100 µL, compared to the incubation in the absence of the toxin and GTPγS (figure 3A). The fact that Dol-P-man as well as the early GPI intermediates were poorly detected in pertussis treated membranes indicates that the transfer of mannose residues into GPI is also stimulated. The data obtained in the presence of pertussis toxin suggest the involvement of the inhibitory regulatory ‘Gαi’ protein, the inhibitory regulatory component of adenylyl cyclase, in the regulation of Dol-P-man synthesis. 3.4. The involvement of regulatory ‘Gs’ in the stimulation of Dol-P-man synthesis The inhibitory effect of GTPγS on the synthesis of Dol-P-man suggest the activation of the Gαi subunit. However, the Dol-P-man synthesis has been also shown to be stimulated via cAMP dependent phosphotyrosine kinase [20]. Therefore, we tested the involvement of ‘Gαs’ GTP binding proteins, the stimulatory regulatory component of adenylyl cyclase. Membranes were pretreated with cholera toxin which is known to activate adenylyl cyclase by stimulating the subunit ‘Gαs’ of the GTP binding protein via ADP-ribosylation, by maintaining it in an activated state. Membranes were first incubated with different concentrations of V. cholera toxin for 10 min then incubated for 30 min, in the presence of GDP[3H]mannose. TLC analysis of the glycolipids extracted from membranes treated with cholera toxin shows a concentration-dependent increase of [3H]mannose incorporation into Dol-P-man and GPIs (figure 4B, C). The incorporation of [3H]mannose is stimulated by a factor of 2 and 3 when membranes were pretreated with 5 µg/100 µL and 10 µg/100 µL, respectively (figure 4D), compared to the incubation in the absence of the cholera toxin. The data suggest a stimulation of Dol-P-man synthase by this toxin, implying that the stimulatory ‘Gαs’ protein is involved in the stimulation and regulation of Dol-P-man synthesis. However the incorporation of [3H]mannose into GPI is limited compared to a higher incorporation of [3H]mannose into Dol-P-man. We speculate that it is due to another mechanism that could regulate the intracellular pool of the mannosylated GPI intermediates. Furthermore, in membranes which were pretreated with 8-bromocAMP, a nonhydrolysable permeable analogue of cAMP,

Figure 3. The effect of GTPγS is abolished by pertussis toxin. Glycolipids were labeled for 30 min via GDP-[3H]mannose in vitro using P. primaurelia membranes pretreated for 10 min with pertussis toxin (C, D) prior to the incubation with 4 µM GTPγS for an additional 10 min before adding the labeling mixture. Glycolipids were extracted as described in figure 1 and analyzed on silica TLC plates using solvent systems A. The top panels show the TLC radioactivity profile of lipids synthesized in vitro using membranes not pretreated with pertussis toxin, preincubated for 10 min without (A) and with 4 µM GTPγS (B).

806

Azzouz et al. [3H]mannose incorporation into Dol-P-man and therefore GPI glycolipid biosynthesis was stimulated by a factor of 1.5 and 3 when membranes were pretreated with 2 µM and 4 µM, respectively (figure 5A). Labeling with GDP[3H]mannose, using membranes pretreated with forskolin, an adenylyl cyclase stimulator, showed a dose-dependent increase of [3H]mannose incorporation into GPI and Dol-P-man over a concentration range from 0.5 µM to 2 µM (figure 5B). In contrast, [3H]mannose incorporation is inhibited in a dose dependent manner when membranes are pre-incubated in the presence of increasing concentrations of [1-(5-isoquinolinesulfonyl)-2-methylpiperazine HCl] (H-7 dihydrochloride) and staurosporine, inhibitors of protein kinases, (figure 5C, D). These data suggest the involvement of ‘Gαs’ in the regulation of Dol-P-man synthesis, through dolichol kinase via the activation of the adenylyl cyclase and cAMP-dependent protein kinases. 4. Discussion

Figure 4. Cholera toxin stimulates GPIs biosynthesis. Glycolipids were labeled for 30 min via GDP-[3H] mannose in vitro using P. primaurelia membranes, pretreated for 10 min with cholera toxin (B, C). Glycolipids were processed as described in figure 1 and analyzed on silica TLC plates using solvent systems A. A. TLC radioactivity profile of lipids synthesized, using untreated membranes. Profiles of three separate experiments were integrated and the amount of radioactivity incorporated in all GPI glycolipids and Dol-P-man was calculated (D), using the Chroma software (Berthold LB 2842).

Heterotrimeric G proteins are not only present at the plasma membranes but also on a variety of intracellular membranes, including Golgi membranes [31], ER membranes [32], and the nucleus [33]. Intracellular G proteins are involved in the regulation of vesicular transport at multiple steps [34–37], such as the transport from ER to Golgi membranes [32] or intra-Golgi transport [38]. We report here that such proteins may be involved in the regulation of the transport of GPI-proteins by regulating the biosynthesis of GPI precursors at the ER. We have used P. primaurelia as a model to study the biosynthesis of glycosylphosphatidyl-inositol anchors [22, 23]. The core glycan of the Paramecium primaurelia polar glycosyl-inositol-phosphoceramides is modified by a mannosyl phosphate unit, linked to the mannose adjacent to glucosamine via a phosphodiester bond. Since the three mannose residues that constitute the GPI core glycan in Paramecium are derived from dolichol phosphate mannose, it was of interest to investigate the biosynthesis of the mannosyl phosphate modification which occurs in two steps: The first step involves the phosphorylation of the Man3-intermediate followed by the mannosylation of the phosphate group via a complex mechanism involving a GalNAc-containing intermediate [23]. Using different nucleotide triphosphates (ATP and GTP) and their corresponding non hydrolyzable analogous (ATPγS and GTPγS), we found that GTPγS inhibits the synthesis of the mannosylated GPIs by inhibiting the synthesis of Dol-P-man. The biosynthesis of the early GPI intermediates, GlcNAc-inositol-phosphoceramide and GlcNinositol-phosphoceramide, is not affected by GTPγS as in the case of mammalian cells [16]. GTPγS inhibits also in vitro the synthesis of GPI glycolipids in the parasitic protozoa Plasmodium falciparum and Trypanosoma brucei (unpublished data) and thus seems to be of broader

Regulation of P. primaurelia glycosylphosphatidyl-inositol biosynthesis

807

Figure 5. Involvement of adenylyl cyclase and cAMP dependent protein kinase. Glycolipids were labeled via GDP-[3H] mannose for 30 min in vitro, using P. primaurelia membranes pretreated for 10 min with 8-bromo-cAMP (A), forskolin (B), H-7-dihydrochloride (C), and staurosporin (D) at different concentrations, as indicated in each figure. Labeled glycolipids were extracted as described in figure 1 and analyzed by TLC. Profiles from three separate experiments were integrated as in figure 5 and the amount of radioactivity incorporated in all GPI glycolipids and Dol-P-man is reported.

significance. Therefore, we investigated the effect of GTPγS in detail on the two enzymes involved in the synthesis of Dol-P-man: dolichol kinase and dolichol phosphate mannose synthase. We show that GTPγS greatly reduces the activity of dolichol kinase but not of Dol-P-man synthase. In the Paramecium system, the observed effect of GTPγS on the synthesis of Dol-P-man suggests that the synthesis of the donor of mannose residues, Dol-P-man, is regulated by heterotrimeric GTP binding proteins involving the adenylyl cyclase/cAMP system and protein phosphorylation. Most of the signal transducing activities of G-proteins are assumed to be associated with the state of activation of their α subunit, which is involved in GDP/GTP exchange and GTP hydrolysis [24, 39]. Certain activated α subunit isoforms can function in stimulatory (αs) or inhibitory (αi) receptorcoupled pathways for regulation of adenylyl cyclase [24].

In the Paramecium system, a stimulation of adenylyl cyclase in vitro leads to the stimulation of Dol-P-man synthesis. This was demonstrated by using cholera toxin indicating that the Dol-P-man synthase is regulated by the stimulatory Gs protein, which is the positive regulator of adenylyl cyclase/cAMP system. The involvement of adenylyl cyclase/cAMP system was further substantiated by using forskolin, an activator of adenylyl cyclase, which stimulates the incorporation of mannose residues into GPI. The involvement of the cAMP-dependent kinases was demonstrated by the use of H-7, DiHCl, a non-specific inhibitor for serine/threonine kinases. However, GTPγS described to mimic the effect of cholera toxin inhibits the synthesis of GPI. This implies that stimulation of adenylyl cyclase initiates a counter regulatory response, which includes the activation of the inhibitory Gi, a negative regulator of adenylyl cyclase/cAMP system [40]. This was

808 shown by the use of pertussis toxin known to inhibit the Gαi subunit. The predominant inhibitory effect could be explained by a difference in concentration between the two classes of G proteins. The events involving the Gi class of protein acting as a negative regulator in the adenylyl cyclase is probably present in much higher concentration than Gs. All data taken together suggest that Dol-P-man synthesis is downregulated by a mechanism that modulates adenylyl cyclase activity by the stimulatory (αs) and the inhibitory (αi) GTP-binding regulatory proteins. The involvement of intracellular G-protein, could be of the stimulatory and/or inhibitory type. Activation by an appropriate physiological signal in vivo could activate Gαs or Gαi to stimulate or inhibit the Dol-P-man synthesis. In vitro both of them could be activated by GTPγS. The synthesis of GPI is a prerequisite not only for the maturation of GPI-proteins in the ER but also for their transport through the secretory pathway. Our work suggests that the transport of GPI-proteins may also be regulated and influenced by compounds that act on the signaling pathways and thus responds to extra cellular events. The physiological in vivo signal is not yet known but could be a degradation product of GPI structures. For instance recent data indicate that the lipid portion as well as the carbohydrate moiety of GPI are involved in intracellular signaling [41, 42].

5. Conclusion In conclusion, we have demonstrated, that one of the possible signaling pathways of GPI biosynthesis is controlled by both Gs and Gi subunits. The former is a positive regulator which stimulates and the latter is a negative regulator, of Dol-P-man synthesis.

Acknowledgments This work was supported by Deutsche Forschungsgemeinschaft (SFB 286), Fonds der Chemischen Industrie, Stiftung P.E. Kempkes and PROCOPE by DAAD/ARNT. H. Shams-Eldin thanks the Wilhelm Schaumann Foundation for doctoral fellowship. The authors thank V. Eckert, E. Peterman, and M. Westermann for their critical reading of the manuscript.

References [1] Ferguson M.A.J., The structure, biosynthesis and functions of glycosylphosphatidyl anchors, and the contributions of trypanosome reseach, J. Cell Sci. 112 (1999) 2799–2809. [2] Nosjean O., Briolay A., Roux B., Mammalian GPI proteins: sorting, membrane residence and functions, Biochim. Biophys. Acta 1331 (1997) 153–186.

Azzouz et al. [3] McConville M.J., Ferguson M.A.J., The structure, biosynthesis and function of glycosylated phosphatidylinositols in the parasitic protozoa and higher eukaryotes, Biochem. J. 294 (1993) 305–324. [4] Field M.C., Menon A.K., Glycolipid anchoring of cell surface proteins, in: Schlesinger M.J. (Ed.), Lipid modification of proteins, CRC-Press, Boca Raton, USA, 1994, pp. 83–134. [5] Stevens V.L., Biosynthesis of glycosylphosphatidylinositol membrane anchors, Biochem. J. 310 (1995) 361–370. [6] Eckert V., Gerold P., Schwarz R.T., GPI-anchors: Structure and Functions, in: Gabius H.J., Gabius S. (Eds.), Glycosciences, Chapman & Hall, Weinheim, Germany, 1997, pp. 223–243. [7] Sharma D.K., Vidugiriene J., Bangs J.D., Menon A.K., A cell-free assay for glycosylphosphatidylinositol anchoring in African trypanosomes. Demonstration of a transamidation reaction mechanism, J. Biol. Chem. 274 (1999) 16479–16486. [8] Milne K.G., Field R.A., Masterson W.J., Cottaz S., Brimacombe J.S., Ferguson M.A.J., Partial purification and characterization of the N-acetylglucosaminyl-phosphatidylinositol de-Nacetylase of glycosylphosphatidylinositol anchor biosynthesis in African trypanosomes, J. Biol. Chem. 269 (1994) 16403–16408. [9] Nakamura N., Inoue N., Watanabe R., Takahashi M., Takeda J., Stevens V.L., Kinoshita T., Expression cloning of PIG-L, a candidate N-acetylglucosaminyl-phosphatidylinositol deacetylase, J. Biol. Chem. 272 (1997) 15834–15840. [10] Mazhari-Tabrizi R., Eckert V., Blank M., Müller R., Mumberg D., Funk M., Schwarz R.T., Cloning and functional expression of glycosyltransferase from parasitic protozoans by heterologous complementation in yeast: the dolichol phosphate mannose synthase from Trypanosoma brucei brucei, Biochem. J. 316 (1996) 853–858. [11] Benghezal M., Benachour A., Rusconi S., Aebi M., Conzelmann A., Yeast Gpi8p is essential for GPI anchor attachment onto proteins, EMBO J. 15 (1996) 6575–6583. [12] Benghezal M., Lipke P.N., Conzelmann A., Identification of six complementation classes involved in the biosynthesis of glycosylphosphatidylinositol anchors in Saccharomyces cerevisiae, J. Cell Biol. 130 (1995) 1333–1344. [13] Leidich S.D., Kostova Z., Latek R.R., Costello L.C., Drapp D.A., Gray W., Fassler J.S., Orlean P., Temperature-sensitive yeast GPI anchoring mutants gpi2 and gpi3 are defective in the synthesis of N-acetylglucosaminyl phosphatidylinositol. Cloning of the GPI2 gene, J. Biol. Chem. 270 (1995) 13029–13035. [14] Schönbachler M., Horvath A., Fassler J., Riezman H., The yeast spt14 gene is homologous to the human PIG-A gene and is required for GPI anchor synthesis, EMBO J. 14 (1995) 1637–1645. [15] Takeda J., Kinoshita T., GPI-anchor biosynthesis, Trends Biochem. Sci. 20 (1995) 367–371. [16] Stevens V.L., Regulation of glycosylphosphatidylinositol biosynthesis by GTP. Stimulation of N-acetylglucosaminephosphatidylinositol deacetylation, J. Biol. Chem. 268 (1993) 9718–9724. [17] Stevens V.L., Zhang H., Kristtyanne E.S., Stimulation of glycosylphosphatidylinositol biosynthesis in mammalian cell-free systems by GTP hydrolysis: evidence for the involvement of membrane fusion, Biochem. J. 341 (1999) 577–584. [18] Konrad M., Merz W.E., Long-term effect of cyclic AMP on N-glycosylation is caused by an increase in the activity of the cis-prenyltransferase, Biochem. J. 316 (1996) 575–581. [19] Konrad M., Merz W.E., Regulation of N-glycosylation, J. Biol. Chem. 269 (1994) 8659–8666. [20] Banerjee D.K., Kousvelari E.E., Baum B.J., cAMP-mediated protein phosphorylation of microsomal membranes increases mannosylphosphodolichol synthase, Proc. Natl. Acad. Sci. USA 84 (1987) 6389–6393. [21] Schubert J., Schmidt R.E., Medof M.E., Regulation of glycosylinositol phospholipid anchor assembly in human lymphocytes, J. Biol. Chem. 268 (1993) 6281–6287.

Regulation of P. primaurelia glycosylphosphatidyl-inositol biosynthesis [22] Azzouz N., Striepen B., Gerold P., Capdeville Y., Schwarz R.T., Glycosylinositol-phosphoceramide in the free living protozoan Paramecium primaurelia. Modification of core glycans by mannosyl phosphate, EMBO. J. 14 (1995) 4422–4433. [23] Azzouz N., Gerold P., Schmidt J., Capdeville Y., Schwarz R.T., Transient N-acetylgalactosaminylation of mannosyl phosphate side chain in Paramecium primaurelia glycosylphosphatidylinositols, Eur. J. Biochem. 267 (2000) 3385–3392. [24] Gilman A.G., G proteins: transducers of receptor-generated signals, Annu. Rev. Biochem. 56 (1987) 615–649. [25] Skorecki K.L., Verkman A.S., Ausiello D.A., Cross talk between stimulatory and inhibitory guanosine 5’-triphosphate binding proteins: role in activation and desensitization of the adenylate cyclase response to vasopressin, Biochemistry 26 (1987) 639–645. [26] Masterson W.J., Doering T.L., Hart G.W., Englund P.T., A novel pathway for glycan assembly: Biosynthesis of the glycosylphosphatidylinositol anchor of trypanosome variant surface glycoprotein, Cell 56 (1989) 793–800. [27] Datema R., Schwarz R.T., Jankowski A.W., Fluoroglucoseinhibition of protein glycosylation in vivo, Eur. J. Biochem. 109 (1980) 331–341. [28] Gerold P., Jung N., Azzouz N., Freiberg N., Kobe S., Schwarz R.T., Biosynthesis of glycosylphosphatidylinositols of Plasmodium falciparum in a cell-free incubation system: inositol acylation is needed for mannosylation of glycosylphosphatidylinositols, Biochem. J. 344 (1999) 731–738. [29] Katada T., Kusakabe K., Oinuma M., Ui M.A., Novel mechanism for the inhibition of adenylate cyclase via inhibitory GTP-binding proteins. Calmodulin-dependent inhibition of the cyclase catalyst by the beta gamma-subunits of GTP-binding proteins, J. Biol. Chem. 262 (1987) 11897–11900. [30] Codina J., Hildebrandt J., Iyengar R., Birnbaumer L., Sekura R.D., Manclark C.R., Pertussis toxin substrate, the putative Ni component of adenylyl cyclases, is an alpha beta heterodimer regulated by guanine nucleotide and magnesium, Proc. Natl. Acad. Sci. USA 80 (1983) 4276–4280. [31] Denker S.P., McCaffery J.M., Palade G.E., Insel P.A., Farquhar M.G., Differential distribution of alpha subunits and beta gamma subunits of heterotrimeric G proteins on Golgi membranes of the exocrine pancreas, J. Cell Biol. 133 (1996) 1027–1040.

809

[32] Schwaninger R., Plutner H., Bockoch G.M., Balch W.E., Multiple GTP-binding proteins regulate vesicular transport from the ER to Golgi membranes, J. Cell Biol. 119 (1992) 1077–1096. [33] Saffitz J.E., Nash J.A., Green K.G., Luke R.A., Ransnas L.A., Insel P.A., Immunoelectron microscopic identification of cytoplasmic and nuclear Gs alpha in S49 lymphoma cells, FASEB. J. 8 (1994) 252–258. [34] Bomsel M., Mostov K., Role of heterotrimeric G proteins in membrane traffic, Mol. Biol. Cell 3 (1992) 1317–1328. [35] Mostov K., Apodaca G., Aroeti B., Okamoto C., Plasma membrane protein sorting in polarized epithelial cells, J. Cell. Biol. 116 (1992) 577–583. [36] Barr F.A., Leyte A., Mollner T., Pfeuffer T., Tooze S.A., Hutter W.B., Trimeric G-proteins of the trans-Golgi network are involved in the formation of constitutive secretory vesicules and immature secretory granules, FEBS. Lett. 294 (1991) 239–243. [37] Melancon P., Glick B.S., Malhotra V., Weidman P.J., Serafini T., Gleason M.L., Orci L., Rothman J.E., Involvement of GTPbinding ‘G’ proteins in transport through the Golgi stack, Cell 51 (1989) 1053–1062. [38] Stow J.L., de Almeida J.B., Narula N., Holtzman E.J., Ercolani L., Ausiello D.A., A heterotrimeric G protein, G alpha i-3, on Golgi membranes regulates the secretion of a heparansulfate proteoglycan in LLC-PK1 epithelial cells, J. Cell Biol. 114 (1991) 1113–1124. [39] Neer E.J., Heterotrimeric G proteins: organizers of transmembrane signals, Cell 80 (1995) 249–257. [40] Wong Y.H., Conklin B.R., Bourne H.R., Gz-mediated hormonal inhibition of cyclic AMP accumulation, Science 255 (1992) 339–342. [41] Schofield L., Tachado S.D., Regulation of host cell function by glycosylphosphatidylinositols of the parasitic protozoa, Immunol. Cell. Biol. 74 (1996) 555–563. [42] Tachado S.D., Gerold P., McConville M.J., Baldwin T., Quilici D., Schwarz R.T., Schofield L., Glycosylphosphatidylinositol toxin of Plasmodium induces nitric oxide synthase expression in macrophages and vascular endothelial cells by a protein tyrosine kinase-dependent and protein kinase C-dependent signaling pathway, J. Immunol. 156 (1996) 1897–1907.