Structural comparisons between the soluble and the GPl

a review see [4, 27]) are high molecular mass proteins. (250-300 kDa) composed of a single polypeptide chain. [14, 24]. It has been shown indirectly ...... sequence of the Paramecium primaurelia G surface protein. A huge protein with a highly ...
692KB taille 3 téléchargements 298 vues
Biol Cell (1992) 75, 217-223 © Elsevier, Paris

217

Original article

Structural comparisons between the soluble and the GPl-anchored forms of the Paramecium temperature-specific 156G surface antigen Nahid Azzouz, Yvonne Capdeville * Centre de Gdndtique Moldculaire, Centre National de la Recherche Scientifique, Associ~ ~t l'Universitd Pierre-et-Marie-Curie, 91198 Gif-sur-Yvette, France (Received 25 May 1992; accepted 27 August 1992) Summary - Biosynthetic labelling experiments performed on Pprimaurelia strain 156, expressing the temperature-specific G surface

antigen, 156G SAg, demonstrated that the purified 156G SAg contained the components characteristic of a GPI-anchor. [3H]ethanolamine, [3H]myo-inositol, [32p]phosphoric acid and [3H]myristic acid could all be incorporated into the surface antigen. Myristic acid labelling was lost after treatment in vitro with Bacillus thuringiensis phosphatidylinositol-specific phospholipase C (PI-PLC). After complete digestion by pronase, a fragment containing the intact GPl-anchor of 156G surface antigen was isolated. This fragment was shown to be hydrophobic and glycosylated and to possess an epitope found specifically in the GPI component of GPl-anchored proteins. The role of the GPI-tail in anchoring the 156G surface antigen into the membrane was assessed by determining that purified 156G molecules with the GPI-anchor could be incorporated into lipid vesicles and red cell ghosts whereas the 156G molecules lacking the GPI-anchor, as result of treatment with B thuringiensis PI-PLC, could not. It has also been shown that the membrane-bound form and the soluble form, obtained after cleavage of the 156G SAg lipid moiety either by an endogenous PI-PLC or by a bacterial PI-PLC, displayed identical circular dichroic spectra. GPl-protein / biosynthetic labelling / phosphatidylinositol-specific phospholipase C / Paramecium

Introduction

Paramecium aurelia, a free living ciliated protozoan, can express a repertoire of different surface antigens (SAgs) which coat, almost entirely, the surface of the cell. These molecules, which undergo antigenic variation triggered by modifying external parameters such as temperature (for a review see [4, 27]) are high molecular mass proteins (250-300 kDa) composed of a single polypeptide chain [14, 24]. It has been shown indirectly [7] that Paramecium SAgs are anchored in the plasma membrane via a glycan phosphatidylinositol (GPI) anchor in the same way as are Trypanosoma variant surface glycoproteins [11] and other eukaryotic surface proteins (for a review see [8, 12, 21]). The lipid anchor of the Paramecium SAgs can be cleaved in vitro by exogenous phosphatidylinositol-specific phospholipases C (PI-PLC), and in vivo by an endogenous phospholipase C like-hydrolase [7]. In both cases, the enzymatic conversion of the m e m b r a n e - b o u n d form of SAg (mSAg) to its soluble form (sSAg) is accompanied by the unmasking of a particular glycosidic immunodeterminant, the cross-reacting determinant (CRD) which was first described in African trypanosomes [3]. The availability of obtaining relatively large amounts of the Paramecium primaurelia 156G SAg, purified in its m e m b r a n e - b o u n d form [1], has opened the way for biochemical and biophysical studies. In the present report, we have characterized biochemi*Correspondence and reprints

cally the G P I anchor of the 156G SAg and we checked its capability to anchor the SAg molecules into membranes. We also describe a procedure for purifying a proteolytic fragment bearing the G P I moiety. This is a prerequisite for completely determining this structure. Finally, we analyzed by circular dichroism (CD) both forms of the 156G SAg, the membrane-bound form and the soluble one, in order to ascertain whether their secondary structures are similar. Indeed, our long-term goal is to understand the topology of two different SAgs expressed together at the cell surface in the course of the antigenic variation process. Thus, the knowledge of the SAg structures and of the types of interactions that occur between these molecules is required. Investigation of the membrane form requires the presence of detergent which complicates the structural studies. The soluble form is much easier to study and would serve as a convenient model if both forms have a similar structure. The present study is an attempt to compare the characteristics of these two forms.

Materials and m e t h o d s

Cells of P. primaurelia strain 156 were grown monoxenicaUy at 24°C, where they all express the G SAg (156G SAg), in hay grass medium inoculated a day before with Enterobacter aerogenes and supplemented with fl-sitosterol (8/zg/ml). The identity of the expressed antigen and the homogeneity of the population were checked by immobilization tests in the presence of specific 156G antibodies [5]. Cultures were used when 10007o of the cells expressed the G SAg.

218

N Azzouz, Y Capdeville

Biosynthetic radiolabelling Before labelling, exponentially grown paramecia were pelleted at 250 g in a GGT centrifuge (Jouan, Paris, France) and washed in Dryl's solution, 1 mM NaH2PO 4, l mM NaHPO 4, 2 mM trisodium citrate, 1.5 mM CaCl 2 (pH 6.8) [9].

[3H]labelling Washed paramecia (2.5 x 105) were incubated in 4 ml of Dryl's solution with [3H]myristic acid-BSA complex (4.6 MBq) for 3 h. Then they were grown in 50 ml of the usual medium for 12 h. For [3H]ethanolamine labelling, washed cells were incubated in Dryl solution with [3H]ethanolamine (7.4 MBq) for 6 h then washed three times before being left in 50 ml of the usual medium for 12 h. For [3H]myo-inositol labelling, washed cells were incubated in Dryl solution in the presence of this radiometabolite (7.4 MBq) plus BSA (1 mg/ml) for 16 h without further growth.

[32p] labelling Paramecia were poorly labelled by direct incubation with [32p]phosphoric acid. Thus, [32Pl-labelling was performed by feeding cells with bacteria previously labelled with [32p] by growth in the Neidhardt medium [22] which was phosphatelimited (0.9 mM) and supplemented with [32P]phosphoric acid (37 MBq). Washed paramecia were mixed with pelleted [32p]-labelled bacteria in an exhausted axenic hay grass medium supplemented with/3-sitosterol. The exhausted medium was obtained as follows: paramecia, grown in the classical medium until exhaustion, were eliminated by centrifugation at 250 g, then the supernatant was centrifuged at 10000 g, autoclaved and kept at 4°C until use.

[35S] labelling In the absence of a defined medium to grow paramecia, [35S] labelling of SAg molecules can be performed by employing a labelling protocol similar to that used for [32P]-labelling. Paramecia were grown in exhausted medium containing /3sitosterol with pelleted [35S]-labelled bacteria. Bacteria were previously grown in minimal medium, sulphate-limited (0.05 mM) and supplemented with [aSS]sulfate (18.5 MBq). After labelling, paramecia were washed three times in Dryl solution before being processed for purification of the 156G SAg. The labelling of the purified 156G mSAg was analyzed by autoradiography, except in the case of biosynthetic labelling by [3H]myo-inositol which was poorly incorporated into paramecia. We therefore checked the incorporation of this component as follows: eight volumes of cold acetone at 8007o were added to different concentrations of the purified [3H]myo-inositol mSAg. The mixture was allowed to stand for 4 h at - 2 0 ° C and was centrifuged at 13 000 g for 30 min. The precipitate was collected and analyzed by scintillation counting along with the corresponding supernatant.

Purification of the membrane-bound form of the SAg Cells were solubilized in a solution of 1070 Triton X-100 (v/v) in the presence of 5 mM p-chloromercuriphenylsulfonic acid OaCMPSA) which inhibits the activity of endogenous phospholipase C-like hydrolase. The purification of 156G mSAg was performed as described previously [1]. Briefy, the Triton X-100 supernatant was dialyzed against 50 mM Tris, 5 mM EDTA (pH 7.4) containing 0.507o Triton X-100 (v/v) and then precipitated by addition of solid ammonium sulfate at 30 and 4007o saturation. The 30 and 4007o pellets containing 156G mSAg were combined, resolubilized in 50 mM Tris, 5mM EDTA (pH 7.4) containing 0.507o Triton X-100 (v/v) and applied to a Biogel A 0.5-m column (Bio-Rad, Paris, France), equilibrated and eluted with 10 mM Tris, 150 mM NaCI, 0.02507e NaN 3 (pH 8) containing 0.2507e DOC. The fractions containing SAg from the Biogel column were concentrated and applied to a TSK 5000 PW size-exclusion column (Beckman, Gagny, France) which was

equilibrated and eluted with 50 mM Tris, 100 mM NaCI, 5 mM EDTA (pH 7.4) containing 1 mM of octaethylene glycol monon-dodecyl ether (Ct2E8) (Nikko chemicals, Tokyo, Japan).

Phosphatidylinositol-specific phospholipase C treatment Purified 156G mSAg labelled with [3H]myristic acid was treated with 2 U/ml of PI-PLC from B thuringiensis for 60 min at 37°C in 50 mM Tris-HC1, 1070Triton X-100 (pH 7.4) and then analyzed by fluorography after electrophoresis on polyacrylamide gels in the presence of SDS.

Protein electrophoresis, autoradiography SDS-PAGE was performed according to Laemmli [19] on 10070 and 5-15070 gradient gels./3-Mercaptoethanol was omitted from the sample buffer for electrophoresis. Gels were stained with Coomassie brilliant blue R-250. Protein concentrations were determined by the BCA protein assay reagent (Pierce, Vitry-surSeine, France). Radiolabelled proteins were visualized, using Kodak-X-OMAT AR films (Kodak, Croissy Beaubourg, France), by fluorography at - 8 0 ° C for [3H], at room temperature for [35S1 or in the presence of a screen at - 8 0 ° C for [32p].

Immunological analysis Immunoblotting experiments were carried out according to Towbin et al [28] with slight modifications [6]. An antiserum raised specifically against the purified membrane-bound form of 156G SAg and an antiserum raised against the purified soluble form of a variant surface glycoprotein of Trypanosoma equiperdum, thus containing antibodies specific for the CRD epitope, were used at a l:100 (v/v) dilution. Peroxidase-labelled antibodies raised against rabbit immunoglobulins (Institut Pasteur, France) were used at a 1:200 (v/v) dilution, with 4-chloro-l-naphtol, as chromogenic substance (Sigma, La Verpilliere, France). Immunogold labelling experiments were performed as previously described [7].

Incorporation of mSAg into red cell ghosts Ghosts were prepared from fresh rabbit erythrocytes according to Falrbanks et al [10]. Pelleted ghosts were resuspended in 50 mM Tris-HC1, I mM MgCl 2, and 20 ~.M ZnCI 2 (pH 7.4) and then mixed with purified [32S]-labelled 156G mSAg and incubated at 37°C for 15, 30 or 60 min. At each time, the sample was centrifuged for 10 min at 20000 g. The pellet was then washed and recentrifuged at 20000 g after which pellets and supernatants were analyzed by scintillation counting.

Incorporation of mSAg into liposomes Egg lecithin (8 mg) in chloroform/methanol (2:1 v/v) was dried to a thin film under a stream of argon. Drying of the lipid film was carried out under vacuum for several hours. The film was then redissolved in 2 ml of 10 mM Tris-HCl, 100 mM NaCI (pH 7.4) containing 1070 Triton X-100. Traces of the purified [35S]labelled 156G mSAg mixed with cold 156G mSAg were added to the dispersed lipid solution at a final lipid-to-protein ratio (w/w) of 100. SM-2 Bio-Beads (Bio-Rad) (0.6 mg), pretreated as described by Holloway [15], were added to the detergent/lipid/protein mixture which was then stirred for 48 h at 4°C. SM-2 Bio-Beads were removed by filtering the solution through glass wool. Protein-free vesicles were prepared using the same protocol but without protein. Purification of the vesicles was performed on a linear sucrose density gradient (5-40°70 w/v) in 10 mM Tris-HCI, 100 mM NaCI (pH 7.5) obtained by centrifugation at 40 000 rev/min in an SW 41 rotor (Beckman), at 4°C for 20 h. 0.7-ml-fractions were collected from the bottom of the 13-ml tubes and their radioactivity and absorbance (550 nm) measured.

GPl-anchor of Paramecium surface antigen

219

Proteolys& of the 156G rnSAg and isolation of its GPl-anchor A kinetic experiment was carried out with pronase at 37°C. 400 ~g of freeze-dried 156 G mSAg previously dialyzed against water were dissolved in 200 t.zl of 50 mM Tris-HCl, 10 mM CaCI 2 (pH 7.4) containing 0.5 mg/ml of pronase (enzyme to substrate ratio of 1:4). At various intervals, an aliquot of the proteolytic reaction mixture corresponding to 15 tzg of 156G mSAg was adjusted to 2 mM PMSF to inhibit the protease activity and kept frozen until analyzed by SDS-PAGE. After complete proteolysis of 2 mg of the mSAg by pronase, the GPI-anchor was extracted with chloroform~methanol~IN H CI (65:35:1, v/v/v) as described by Kennington et al [18]. The organic phase was dried under vacuum, and then resuspended in water containing 0.1070 TFA. The sample was then injected, at room temperature, onto a reverse phase column (ODS-C~8) (Beckman), and eluted using a gradient of 0-100070 acetonitrile, 0.1% TFA, with a Beckman Model 344 chromatography system equipped with a model 160 absorbance detector and a 214 nm filter.

Circular dichrohsm (CD)

mSAg

~

1

2

3

4

Fig 1. [~H]ethanolamine, [3H]myristic acid and [32P]phosphoric acid incorporation into the 156G mSAg. The 156G mSAg purified from cells which were biosynthetically labelled with these components was analyzed by SDS-PAGE. Lanes: 1, Coomassie blue stained gel; 2, autoradiogram of the 156G mSAg labelled with [~H]ethanolamine; 3, autoradiogram of the 156G mSAg labelled with [3H]myristic acid; 4, autoradiogram of the 156G mSAg labelled with [3-'P1.

CD was measured with a Jobin-Yvon Mark V model instrument (Jobin-Yvon, Longjumeau, France). In the 195-260 spectral zone the measurements were made in cuvettes with optical path of 0.2 cm at 20"C. The value of 110 was used for the mean residue weight.

Reagents Pronase, leupeptin, phenyl-methylsulphonyl fluoride (PMSF) and p-chloromercuriphenylsulfonic acid (pCMPSA) were purchased from Sigma; octaethylene glycol mono-n-dodecyl ether (C12E8) was from Nikko chemicals (Tokyo, Japan), Bacillus thuringiensis phosphatidylinositol-specific phospholipase C was from Immunotech (Marseille, France), 9,10 (n)-[~H]myristic acid (SA: 1.5-2.2 TBq/mmol), I-[~H] ethan-l-ol-2-amine hydrochloride (SA: 0.18-1.1 TBq/mmol), myo-2-[~H]inositol (SA: 370-740 GBq/mmol), [32P]phosphoric acid (SA: 7.4 GBq/mmol) and [35S]sulfate (Sulphur-35; SA: 0.9-1.5 TBq/mg) were from Amersham (Les Ulis, France).

Results Characterization of the 156G anchor by biosynthetic labelling Following biosynthetic labelling successively with [~H]myristic acid, [32p]phosphoric acid, [~H]ethanolamine and [3H}myo-inositol, the Paramecium 156G mSAg was purified and the extent of incorporation o f each of these radiolabelled metabolites was assessed. The purified 156G mSAg was found to be labelled with [3H]myristic acid, [32P]phosphoric acid, and [3H]ethanolamine (fig 1). The labelling by [3H]myristic acid was lost after treatment of the purified 156G mSAg by B thuringiensis P I - P L C , concomitantly the CRD epitope was unmasked and the electrophoretical mobility of the SAg was decreased (fig 2). Thus, all the effects typical of the action of a P I - P L C on a GPI-anchored protein [12] are observed here. As ['H]myo-inositol was only very weakly incorporated into paramecia, the SAg molecules were poorly labelled. Thus, to ascertain the presence of myo-inositol in the structure of the 156G SAg, different quantities of purified [3H]myo-inositol-labelled 156G SAg were precipitated by acetone at - 2 0 ° C and analyzed by scintillation counting. The number of counts increased linearly with the quanti-

ty of 156G mSAg precipitated by cold acetone (not shown) demonstrating the presence of [3H]myo-inositol.

Isolation of the 156G anchor To determine the conditions for complete digestion of the 156G mSAg, a time course experiment was carried out with pronase used at an enzyme-to-substrate ratio of 1:4. The proteolytic digestion mixture was analyzed by SDS-PAGE (fig 3). Four proteolytic fragments were detected early during the course of proteolysis (lane 3). These fragments appeared not to be degraded further during the remainder of the experiment (lanes 3 - 1 0 ) . The apparent MrS of these stable fragments on SDS-PAGE were between 20000 and 40000. Intact 156G mSAg was no longer observed at 24 h (lane 10). The GPI-anchor fragment was extracted into an organic phase after 24 h of digestion by pronase. The organic phase was then passed over a C i 8 0 D S column and the fragment, indicated by an arrow, was eluted at about 7007o acetonitrile (fig 4A). The analysis of this fragment by SDSP A G E shows an apparent M r of about 10000. It was stained by Coomassie blue and it was periodic acid, Schiff's (PAS) reagent positive (results not shown). After treatment by B thuringiensis P I - P L C , this fragment displayed the CRD epitope (fig 4B). Thus, it can be concluded that this fragment, which is hydrophobic, glycosylated and contains the CRD epitope, bears the G P I - a n c h o r of the 156G mSAg.

Role of the GPI structure in membrane anchorage The role of the G P I in anchoring the SAg to the membrane was investigated by studying the incorporation into red cell ghosts and the reconstitution into vesicles using the purified 156G SAg either in its membrane-bound form or in its soluble f o r m obtained after B thuringiensis PIPLC treatment. In order to follow the incorporation of the 156G SAg, which has a high content of cysteine residues (10°70), paramecia were previously biosynthetically labelled with [35S].

220

N Azzouz, Y Capdeville

A

B

C "+':";: +

I

"

mSAg

I

Pl-PLC

1

2

l

2

Fig 2. B thuringiensis PI-PLC treatment of the 156G mSAg. After biosynthetic labelling by [3H]myristic acid, the purified 156G mSAg was treated by B thuringiensis PI+PLC which cleaves off the diaeyglycerol part of GPI-anchored proteins thus unmasking the CRD epitope and then analyzed by SDS-PAGE and immunoblotting. A. Protein staining with Coomassie brilliant blue. B. Autoradiogram corresponding to A. C. Immunoblotting using anti-CRD antibodies. Lane l, purified 156G mSAg; lane 2, 156G sSAg derived from the 156G mSAg by enzymatic conversion performed in vitro with B thuringiensis PI-PLC.

Molecular mass (KDa) mSAg ~

"i

II-"+,

~

~

"--

- -

-94 L_

- -

----

-43

-30 2, 7 ~

"7

-20

• "/d:

I

2

3

4

5

6

7

8

9

I0

Fig 3. Time course of hydrolysis by pronase of the 156G mSAg. Pronase (in 50 mM Tris-HCl, 10 mM CaCI2, pbl 7.4) was added to freeze-dried 156G mSAg (see Materials and methods). The digestion was performed at 37°C. Samples corresponding to 15 ~g of 156G mSAg were taken and analyzed by 10070SDS-PAGE at the following times: 15 min, 1 h, 2 h, 3 h, 4 h, 5 h, 6 h, 24 h, lanes 3-10, respectively. Lanes: 1, analysis of 10/zg of pronase; 2, purified 156G mSAg (control).

Incorporation of the 156G mSAg into red cell ghosts

Incorporation of the 156G mSAg into liposomes

Purified [35S] 156G mSAg was added to red cell ghosts. At different times (15, 30 and 60 min), samples containing identical quantities of [35S] 156G mSAg were taken and centrifuged at 20000 g. The radioactivity of the pellets and the corresponding supernatants were determined. The incorporation of the [35S] 156G mSAg was about 50°7o at the first point and did not increase with time. No incorporation was obtained when [3ss] 156G mSAg molecules were first treated by a bacterial PI-PLC.

We attempted to incorporate purified [3sS]-labelled 156G mSAg molecules into lipid vesicles made of egg lecithin (see Materials and methods), using three different protocols. The efficiency of incorporation was assessed by scintillation counting of the pelleted vesicles. For the first protocol used, the dialysis method, we tested three detergents, Triton X-100, C~2E8, and n-OG to solubilize lipids and to resuspend the freeze-dried 156G SAg. In the second protocol, the dilution method [17], we

GPl-anchor of Paramecium surface antigen

221 m

A b'3

o

Bc~