EXPERlMENTAL PARASITOLOGY 74, 452462 (1992)
Plasmodium falciparum and Plasmodium chabaudi: Characterization of Glycosylphosphatidylinositol-Degrading Activities CATHERINE
BRAUN-BRETON, *J THIERRY BLISNICK,* PATRICIA BARBOT,* LUIZ PEREIRA DA SILVA,* AND GORDON LANGSLEY*
ROLAND Bihow,t3* *Unit
of Experimental
du Dr.
Roux,
Parasitology, URA CNRS 146, Department Paris 75015, France; and tMax Planck Institut
of Immunology, fiir Biologie,
lnstitut Tiibingen,
Pasteur, Germany
25 rue
BRAUN-BRETON, C., BLISNICK, T., BARBOT, P., Bij~ow, R., PEREIRA DA SILVA, L., AND LANGSLEY, G. 1992. Plasmodium falciparum and Plasmodium chabaudi: Characterization of glycosylphosphatidylinositol-degrading activities. Experimental Parasitology 74, 452462. Merozoites of malaria parasites have a membrane-bound serine protease whose solubilization and subsequent activity depend on a parasite-derived glycosylphosphatidylinositol-phospholipase C (GPI-PLC). The GPI-degrading activities from both Plasmodium falciparum and Plasmodium chabaudi have been characterized and partially purified by phenylboronate chromatography. They are membrane-bound, developmentally regulated, calcium-independent enzymes and as such they resemble GPI-PLC of Trypanosoma brucei. Furthermore, a T. brucei GPI-PLC-specific monoclonal antibody (mAT3) immunoprecipitates the plasmodial GPI-degrading activity. Thin-layer chromatography is suggestive of two activities: a GPI-PLC and a phospholipase A. o 1992 Academic PESS, hc. INDEX DESCRIPTORS AND ABBREVIATIONS: Plasmodia; Phospholipase; Glycosylphosphatidylinositol; Cross-reacting determinant (CRD); Diacylglycerol (DAG); Glycosylphosphatidylinositol (GPI); Glycosylphosphatidylinositol-specific phospholipase C (GPI-PLC); Phosphatidylinositol-specific phospholipase C (PI-PLC); Phosphatidylinositol (PI); Phospholipase A (PLA); Red blood cell (RBC); Variant surface glycoprotein (VSG); Tritiated membrane-form variant surface glycoprotein ([3H]mfVSG).
During the past few years, a new type of membrane anchor has been described for a wide variety of eukaryotic membrane proteins (Low and Saltiel 1988). This anchorage involves a covalent linkage of the COOH terminus of the protein to a glycosylphosphatidylinositol domain (GPI), (Ferguson et al. 1988). The membrane form of GPI-anchored proteins can be converted to a soluble form by digestion with serum phospholipase D or with Ca*+-independent phosphatidylinositol-specific phospholipase C (PI-PLC). The latter reaction liberates diacylglycerol (DAG), which has been ’ To whom correspondence should be addressed. Fax: (331) 42 73 22 40. * Present address: Department of Medical Microbiology, Sherman Fairchild Science Building, Stanford, California, U.S.A. 452 0014-4894192 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
reported to act as a second messenger (for review see Berridge 1987). Several Ca2+-independent PI-PLC have been described and, among these, two glycosylphosphatidylinositol-specific phospholipases (GPI-PLC) have been purified: one from Trypanosoma brucei and the other from rat liver plasma membranes (Biilow and Overath 1986; Hereld et al. 1986; Fox et al. 1986, 1987). The biological role of these enzymes has also been investigated. In rat liver cells, the production of DAG was correlated with the activation of GPIPLC following binding of insulin to the insulin receptor (Saltiel et al. 1988). In trypanosomes, it has been proposed that the GPI-PLC is responsible for the rapid release of the variant surface glycoprotein (VSG) from the cell surface (Turner et al. 1985). No direct evidence for this biological role, however, has been obtained and acti-
CHARACTERIZATION
OF
PLASMODIA
vation of T. brucei GPI-PLC and/or solubilization of VSG were detected only as a result of osmotic lysis of trypanosomes (Cardoso de Almeida and Turner 1983). The trypanosomal enzyme (37 to 40 kDa) is membrane bound and probably located in submembrane vesicles (Biilow and Overath 1986; Hereld et al. 1986; Fox et al. 1986; Grab et al. 1987; Overath et al. 1987). Free phosphatidylinositol is a poor substrate for the enzyme, but it solubilizes several different GPI-anchored proteins (Btilow and Overath 1986; Hereld et al. 1986; Fox et al. 1986; Overath et al. 1987). The cDNA coding for the T. brucei GPI-PLC has been recently cloned and sequenced (Hereld et al. 1988a; Carrington et al. 1989). The deduced mass of 40,760 Da corresponds to the apparent mass of the active lipase. Finally, no homology between the cDNA sequence of the T. brucei GPI-PLC and the nucleotide sequences described for other known phospholipases C was observed (Hereld et al. 1988a; Carrington et al. 1989). We have recently described a P. falciparum protein, ~76, which exhibits properties of a serine protease (Braun-Breton et al. 1988). The membrane form of p76 is anchored to the rhoptry membrane via a GPI domain and does not exhibit proteolytic activity. However, following cleavage by exogenous PI-PLC (Bacillus cereus, Stuphylococcus aureus, and T. brucei enzymes were used), the PI-PLC-solubilized form of p76 exhibits a detectable serine protease activity (Braun-Breton et al. 1988). In merozoites, solubilization of p76 results in the exhibition of the cross-reacting determinant (CRD) (Braun-Breton et al. 1990). The CRD epitope is present only on PI-PLCsolubilized GPI-anchored proteins. In vivo, solubilization of p76 is necessary to reveal its proteolytic activity (Braun-Breton and Pereira da Silva 1988). These results suggested the presence of a compartmentalized, developmentally regulated PI-PLC in the parasite. This type of enzyme regulation may be a feature of all plasmodia, as
453
GPI-PHOSPHOLIPASE
we have recently identified in the rodent malaria parasite Plasmodium chabaudi a similar serine protease activity that also depends on cleavage by PI-PLC (BraunBreton et al. submitted for publication). In this paper, we report the characterization of a developmentally regulated, Ca*‘independent, GPI-degrading activity from both P. falciparum and P. chabaudi. This activity resolves into a phospholipase C (PLC) and a phospholipase A (PLA). The GPI-PLC could be responsible for the in vivo induction of the previously described serine protease activity which correlates with merozoite maturation and red blood cell invasion. The biochemical characterization of these enzymes and eventual cloning of the corresponding genes should aid in the clarification of their physiological role. MATERIALS Strains chabaudi
AND
METHODS
and parasite production. Clone F of the P. IP-IPCl strain was injected into 2-month-old BALB/c mice. The strain characteristics and
female the methods used for its culture and synchronization have been previously described (Falanga et al. 1984). Red blood cells (RBC) were recovered from infected mice when the parasitemia was 50% and osmotically lysed as described (Braun-Breton et al. 1986). Synchronous cultivation of the P. falciparum Palto Alto FUP-Uganda strain was performed (Braun-Breton et al. 1988) and trophozoite-infected RBC were purified on a Percoll-sorbitol gradient prior to further cultivation. The maturation of schizonts was followed by Giemsa staining and the concentrated schizonts were lysed by osmotic shock (Braun-Breton et al. 1986). Preparation of protein extracts. Parasite proteins were extracted by incubation of osmotically lysed cells in 1 vol of 2% NP-40, 100 mMTris, pH 8.0, and 10 mM EDTA for 20 min at room temperature. Preparation of [9,10(n)-3H]myristate-labeled mflSG. Trypanosomes of the Botat-1 variant of Trypanosoma equiperdum (Capbem et al. 1977) were pu-
rified from the blood of C3H-infected female mice (Hereld et al. 1988b). [‘H]Myristate labeling of trypanosomes and isolation of the membrane-form variant surface glycoprotein (mfVSG) were performed following the protocol described by Hereld et al. (1988b). The [3H]mfVSG preparation (12 mg/ml) was stored at -80°C. GPI-anchor-degrading
activity assay. GPIdegrading activity was assayed by its ability to cleave
454
BBAUN-BRETON
[3H]mfVSG, generating butanol-extractable 3H materials and unlabeled soluble VSG. Protease-free PIPLC of different origins were tested. These were purified B. cereus PI-PLC (Boehringer) and purified T. brucei GPI-PLC (a kind gift of Dr. P. Overath). NP-40 extracts from P. fulciparumand P. chabaudi-infected red blood cells were assayed. Purified 13H]myristic acid-labeled mfVSG (2 ug = 40 pmol, 7 x lo3 cpmug) was incubated with PI-PLC or parasite extract in 1% NP-40, 50 mM Tris, pH 8.0, 5 mM EDTA, for 30 min at 37°C. The released [3H]DAG was extracted with 500 ul of butanol (water saturated). The radioactivity in 400 pl of the organic phase was determined by liquid scintillation counting. When required the Triton X-l 14 partition of P. chabaudi proteins was performed as described (Bordier 1988). Phosphatidylinositol (PZ)-degrading activity. The activity was assayed by its ability to cleave L-3phosphatidyl-[2-3H]inositol, generating water-soluble 13H]inositol, butanol-extractable product, and uncleaved [3H]PI. [3H]PI (10 uM solution, 10-15 Cii mmol) was diluted 1:300 in 1% NP-40,50 mM Tris, pH 8.0, 5 mM EDTA (60,000 cpm per assay) and incubated for 30 min at 37°C with NP-40 extracts from P. falciparum and P. chubaudi merozoites, together with the different PI-PLC preparations. These were B. cereus PI-PLC (from Boehringer), B. thuringiensis PIPLC (from Immunotech), T. brucei PI-PLC (from Dr. P. Overath), and phospholipase A, (from Sigma). Butanol extraction was then performed and the radioactivity was determined by liquid scintillation counting of the aqueous and organic phases. Thin-layer chromatography. The products released from [3H]mfVSG in the butanol phase (400 ul), by Listeriu monocytogenes PI-PLC and heat-inactivated (2 hr at 60°C) human serum (phospholipase D) taken as controls as well as P. chabaudi merozoite and P. falciparum blood stage extracts were analyzed by thinlayer chromatography. After evaporation of the solvent, the extracted material was redissolved in 40 ul of CHCl&H,OH/H,O (10:10:3) and chromatographed on a thin-layer Silica Gel 60 (Merck; heated 1 hr at 100°C prior to use) using CHCl,/CH,OH/H,O (65:25:4) as solvent. The plate was sprayed with EN3HANCE (New England Nuclear) and fluorographed. Commercially available sn-1-stearoyl-2-[i4C]arachidonylglycerol, L-3-phosphatidyl-[2-‘Hhnositol and [9,10(n)‘Hlmyristic acid (Amersham) were used as controls for migration of DAG, PI, and myristic acid. Phenylboronate column chromatography of GPZdegrading activity. Detergent extracts of P. fulciparum or P. chabuudi proteins were applied to a phe-
nylboronate (Sigma) column (10 vol beads/l vol extract). The bound proteins were eluted by subsequent addition of 10 mM sorbitol, 100 mM sorbitol, 1 M guanidine, and 3 M guanidine. Starting with 100 mg of parasite proteins, about 10 mg bound to the phenylboronate column and 1 mg was recovered in the 10 mM
ETAL. sorbitol eluate, 2 mg in the 100 mM sorbitol eluate, 2 mg in the 1 M guanidine eluate, and 4 mg in the 3 M guanidine eluate. The different eluates were concentrated to the volume of the extract applied to the column and dialyzed for 48 hr at 4°C against 1% NP-40,50 mM Tris, pH 8.0, 5 mM EDTA in the presence of 50 &ml leupeptin and pepstatin. Zmmunoprecipitation activity. Twenty to
of the GPZ-anchor-degrading
40 ul of detergent-extracted parasite proteins was incubated with 10 to 20 ul of Maps (Bio-Rad kit)-purified, protease-free IgG in 50 mM Tris, pH 8.0, 5 mM EDTA, and 1% NP-40 for 1 hr at 37°C. After the addition of 200 ul of 10% IgSorb (The Enzyme Center, Inc., Malden, MA 02148) and incubation for a further 30 min at 4°C the suspension was centrifuged. The supematant was assayed for GPIanchor-degrading activity. Western blot analysis. Western blots of phenylboronate column eluates from P. fulcipurum and P. chabaudi were performed using a monoclonal antibody, mAT3, raised against T. brucei GPI-PLC (Btilow and Overath 1986). The antigens were visualized after incubation with anti-mouse alkaline phosphataselabeled immunoglobulins and reaction with nitro blue tetrazolium substrate (Promega Biotec). RESULTS
To characterize the PI-PLC-like activity present in P. falciparum and P. chabaudi parasites, we used an enzyme assay similar to that described for T. brucei (Hereld et al. 198813). The assay estimates the ability of 1% NP-40 extracts of P. falciparum and P. chabaudi to liberate [3H]myristate-labeled DAG from mfVSG of T. equiperdum. As shown in Table I the GPI-degrading activity was first detected at the end of the intraerythrocytic cycle and increased as young schizonts (less than four nuclei) matured into merozoites. The P. falciparum activity was essentially recovered in the sedimentable fraction and not in the soluble fraction (Table II). This suggests that the activity is membrane associated. The extensive washing of the membrane fraction to avoid contamination by soluble components is probably responsible for the loss of activity during the separation. In phase separation experiments with Triton X-114, the P. chabaudi GPI-degrading activity behaved as if it were membrane bound (Table II). Taken together these results suggest
CHARACTERIZATION
OF PLASMODIA
Time Course of GPI-Anchor-Degrading
TABLE I Activity in P. falciparum
Protein concentration
Source of enzyme VSG lipase
GPI-PHOSPHOLIPASE
Activity (pmoVml/min)
15 p&ml
455
and P. chabaudi
Specific activity (pmol/mg/min)
3750
2.5 x 105
Activity relative to schizonts P-N)
P. falciparum
NP-40 extracts Mature trophozoites Schizonts, ~4 nuclei Schizonts, 4-8 nuclei Schizonts, 38 nuclei Merozoites
5 5 5 5 2
mglml mg/ml mg/ml mg/ml mg/ml
25 150 500 750 600
5 30 100 150 300
0.05 0.3 1 1.5 3
5 5 5 5
mg/ml mg/ml mg/ml mg/ml
100 500 1000 1500
20 100 200 300
0.2 1 2 3
P. chabaudi
NP-40 extracts Schizonts, ~4 nuclei Schizonts, 4-8 nuclei Schizonts, 28 nuclei Merozoites
Note. The activities presented correspond to the average values of at least three independent experiments.
that the P. falciparum and P. chabaudi phospholipases are membrane-bound proteins. Next, using the same assay, we determined the effects of cations, thiol reducing TABLE II Relative GPI-Degrading Activity in Soluble and Membrane Fractions
Fraction
Relative activity (%I
P. falciparum
NP-40 Soluble Membrane associated
100 co.05 40-50
P. chabaudi
NP-40 Aqueous phase Triton phase
100 co.05 SO-90
Note. P. falciparum parasites were lysed by osmotic shock and soluble and membrane fractions were separated by differential centrifugation (Braun-Breton et al. 1986) or by direct extraction of proteins with 1% NP-40. The relative activity of each fraction was compared. For P. chabaudi osmotically lysed parasites, the comparison was made between a Triton X-l 14 partition of soluble and membrane components and a direct extraction by 1% NP-40.
agents, detergents, and protease inhibitors on the P. falciparum and P. chabaudi phospholipases. These effects were compared to those obtained with T. brucei GPI-PLC and B. cereus PI-PLC in the same enzyme assay (Table III). The plasmodia phospholipases are similar to T. brucei GPI-PLC in a number of characteristics. First, Ca*’ and EDTA had no effect on their activity, indicating that they are calcium-independent, GPI-degrading activities. Second, the activities exhibited an acute sensitivity to Zn*+ ions and to pCMPS and a marked stimulation by DTT, characteristics which are consistent with the requirement for a reduced sulfhydryl group for maximal activity. Third, the liberation of [3H]myristatelabeled fatty acid was sensitive to sodium deoxycholate and slightly sensitive to Triton X-100. Finally, the activity of the different phospholipases was unaltered by the presence of the protease inhibitors leupeptin and pepstatin. The T. brucei phospholipase C appears to be specific for GPI as it cleaves efficiently the VSG glycolipid anchor, but cleaves PI
456
BRAUN-BRETON
ET AL.
TABLE III The Effects of Various Reagents on the GPI-Anchor-Degrading
Activities of Different Phospholipases
% GPI-degrading P. falciparum
P. chabaudi
activity T. brucei
B. cereus
None
100
loo
100
100
CaCI, 2mM 10 mM
100 80
100 90
100 95
nt 90
80
nt
100
loo
85 95
nt 100
EDTA 2mM 10 mM
pCMPS 5mM
I
10
20
100
ZnCI, 2mM IO mM
50 30
nt 2
90 0
100 2
DTT 2mM 10 mM
140 200
130 150
170 205
loo 100
Triton X-100 0.2% 1%
100 66
nt 40
95 70
100 70
15 2
25 2
15 2
85 70
100
100
100
100
Nadeoxycholate 0.2% 1% Leupeptin pepstatin
P. falciparum and P. chabaudi GPI-degrading activities were measured in 1% NP-40 extracts. The T. GPI-PLC (2.5 x lo-* U/ml) and B. cereus PI-PLC (2.5 x lo-* U/ml) were purified enzymes. The numbers represent the activity as a percentage of controls and correspond to the average value of three independent experiments. Note. brucei
to a much lesser extent (Btilow and Overath 1986; Hereld et al. 1986; Fox et al. 1986). The P. fulciparum extract contains a phospholipase C activity, since the p76 serine protease exhibits a CRD epitope after its cleavage by this enzyme (Braun-Breton et al. 1990). The products of [3H]mfVSG degradation by a P. chubaudi merozoites extract and a P. fulcipurum extract were analyzed by thin-layer chromatography. Both plasmodial enzymes and the L. monocytogenes PI-PLC (Mengaud et al. 1991) generate products which comigrate with control diacylglycerol (Fig. 1). These results confirm the presence of a GPI-PLC in P. fulciput-urn and P. chubuudi extracts. A second
and major product, comigrating with myristic acid, is generated by the plasmodial extracts. The enzyme(s) responsible for the generation of myristic acid has not been identified but should be a PLA activity. Depletion of the plasmodial GPI-degrading activity was obtained with different inhibitors active on various PI-PLCs but not reported to be active on known PLAs; depletion of the activity was also obtained by immunoprecipitation with antibodies to T. brucei GPI-PLC (Table VI). This suggests that the generation of myristic acid might be the consequence of further degradation (by the PLA activity) of the diacylglycerol generated by the parasite PLC. We cannot ex-
CHARACTERIZATION
OF
PLASMODIA
DAG
myrtsttc
acid
+
3
origin
j
1. Products of the P. chabaudi and P. falciGPI-anchor-degrading activity. Thin-layer chromatography of butanol-extracted material released from [3H]mfVSG was performed on Silica Gel 60 plate as described (under Materials and Methods). Lane 1, sn-1-stearoyl-2-[r4C]arachidonylglycerol used as a control for migration of DAG; lane 2, L-3phosphatidyl-[2-3H]inositol; lane 3, [9,10(n)-3H]myristic acid; lane 4, material released by a P. chabaudi merozoites extract; lane 5, material released by a P. fulciparum blood stage extract; lane 6, material released by human serum phospholipase D (2 hr at 60°C heat-inactivated human serum); lane 7, material released by control L. monocytogenes LO28 culture supematant; lane 8, same as lane 7 with a deposit twice that in lane 7.
FIG. parum
elude, however, that the parasite PLA activity is Ca2+ -independent, membrane associated, sensitive to thiol reagents and detergents, and immunoprecipitated by anSerum tibodies to T. brucei GPI-PLC. phospholipase D produces a different species, presumably phosphatidic acid, migrating close to the loading point of the samples (Davitz et al. 1987). In some extracts, contamination with phospholipase D (probably from host serum) was detected (data not shown).
451
GPI-PHOSPHOLIPASE
We have also investigated the possibility that the plasmodial phospholipases, like the T. brucei enzyme, are GPI specific. L3H]mfVSG and phosphatidyl-[3H]inositol were compared as substrates for different phospholipases. Phospholipase C digestion of both substrates would result in the release of radioactive material which would partition differently from the substrate: [3H]inositolphosphate into the aqueous phase, versus phosphatidyl-[3H]inositol into the organic phase and likewise [3H]DAG into the organic phase, versus [3H]mfVSG into the aqueous phase. As shown in Table IV, bacterial PI-PLC are able to digest both substrates. In contrast, PI is a poor substrate for both the trypanosomal and malarial enzymes. Phospholipase A2 was used as a control in these experiments. Phospholipase A2 digestion of phosphatidyl[3H]inositol yields myristic acid and monoacylated phosphatidyl-[3H]inositol, both partitioning in the organic phase like the uncleaved PI. As expected, no release TABLE
IV
f3HlPI cpm Substrate phospholipase
released in the aqueous phase
[3H1mfVSG cpm released in the organic phase
2300-4020
28X1-7000
5500-5960
4OOOA500
Plasmodium falciparum and Plasmodium chabaudi: Characterization of Glycosylphosphatidylinositol-Degrading Activities CATHERINE
BRAUN-BRETON, *J THIERRY BLISNICK,* PATRICIA BARBOT,* LUIZ PEREIRA DA SILVA,* AND GORDON LANGSLEY*
ROLAND Bihow,t3* *Unit
of Experimental
du Dr.
Roux,
Parasitology, URA CNRS 146, Department Paris 75015, France; and tMax Planck Institut
of Immunology, fiir Biologie,
lnstitut Tiibingen,
Pasteur, Germany
25 rue
BRAUN-BRETON, C., BLISNICK, T., BARBOT, P., Bij~ow, R., PEREIRA DA SILVA, L., AND LANGSLEY, G. 1992. Plasmodium falciparum and Plasmodium chabaudi: Characterization of glycosylphosphatidylinositol-degrading activities. Experimental Parasitology 74, 452462. Merozoites of malaria parasites have a membrane-bound serine protease whose solubilization and subsequent activity depend on a parasite-derived glycosylphosphatidylinositol-phospholipase C (GPI-PLC). The GPI-degrading activities from both Plasmodium falciparum and Plasmodium chabaudi have been characterized and partially purified by phenylboronate chromatography. They are membrane-bound, developmentally regulated, calcium-independent enzymes and as such they resemble GPI-PLC of Trypanosoma brucei. Furthermore, a T. brucei GPI-PLC-specific monoclonal antibody (mAT3) immunoprecipitates the plasmodial GPI-degrading activity. Thin-layer chromatography is suggestive of two activities: a GPI-PLC and a phospholipase A. o 1992 Academic PESS, hc. INDEX DESCRIPTORS AND ABBREVIATIONS: Plasmodia; Phospholipase; Glycosylphosphatidylinositol; Cross-reacting determinant (CRD); Diacylglycerol (DAG); Glycosylphosphatidylinositol (GPI); Glycosylphosphatidylinositol-specific phospholipase C (GPI-PLC); Phosphatidylinositol-specific phospholipase C (PI-PLC); Phosphatidylinositol (PI); Phospholipase A (PLA); Red blood cell (RBC); Variant surface glycoprotein (VSG); Tritiated membrane-form variant surface glycoprotein ([3H]mfVSG).
During the past few years, a new type of membrane anchor has been described for a wide variety of eukaryotic membrane proteins (Low and Saltiel 1988). This anchorage involves a covalent linkage of the COOH terminus of the protein to a glycosylphosphatidylinositol domain (GPI), (Ferguson et al. 1988). The membrane form of GPI-anchored proteins can be converted to a soluble form by digestion with serum phospholipase D or with Ca*+-independent phosphatidylinositol-specific phospholipase C (PI-PLC). The latter reaction liberates diacylglycerol (DAG), which has been ’ To whom correspondence should be addressed. Fax: (331) 42 73 22 40. * Present address: Department of Medical Microbiology, Sherman Fairchild Science Building, Stanford, California, U.S.A. 452 0014-4894192 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
reported to act as a second messenger (for review see Berridge 1987). Several Ca2+-independent PI-PLC have been described and, among these, two glycosylphosphatidylinositol-specific phospholipases (GPI-PLC) have been purified: one from Trypanosoma brucei and the other from rat liver plasma membranes (Biilow and Overath 1986; Hereld et al. 1986; Fox et al. 1986, 1987). The biological role of these enzymes has also been investigated. In rat liver cells, the production of DAG was correlated with the activation of GPIPLC following binding of insulin to the insulin receptor (Saltiel et al. 1988). In trypanosomes, it has been proposed that the GPI-PLC is responsible for the rapid release of the variant surface glycoprotein (VSG) from the cell surface (Turner et al. 1985). No direct evidence for this biological role, however, has been obtained and acti-
CHARACTERIZATION
OF
PLASMODIA
vation of T. brucei GPI-PLC and/or solubilization of VSG were detected only as a result of osmotic lysis of trypanosomes (Cardoso de Almeida and Turner 1983). The trypanosomal enzyme (37 to 40 kDa) is membrane bound and probably located in submembrane vesicles (Biilow and Overath 1986; Hereld et al. 1986; Fox et al. 1986; Grab et al. 1987; Overath et al. 1987). Free phosphatidylinositol is a poor substrate for the enzyme, but it solubilizes several different GPI-anchored proteins (Btilow and Overath 1986; Hereld et al. 1986; Fox et al. 1986; Overath et al. 1987). The cDNA coding for the T. brucei GPI-PLC has been recently cloned and sequenced (Hereld et al. 1988a; Carrington et al. 1989). The deduced mass of 40,760 Da corresponds to the apparent mass of the active lipase. Finally, no homology between the cDNA sequence of the T. brucei GPI-PLC and the nucleotide sequences described for other known phospholipases C was observed (Hereld et al. 1988a; Carrington et al. 1989). We have recently described a P. falciparum protein, ~76, which exhibits properties of a serine protease (Braun-Breton et al. 1988). The membrane form of p76 is anchored to the rhoptry membrane via a GPI domain and does not exhibit proteolytic activity. However, following cleavage by exogenous PI-PLC (Bacillus cereus, Stuphylococcus aureus, and T. brucei enzymes were used), the PI-PLC-solubilized form of p76 exhibits a detectable serine protease activity (Braun-Breton et al. 1988). In merozoites, solubilization of p76 results in the exhibition of the cross-reacting determinant (CRD) (Braun-Breton et al. 1990). The CRD epitope is present only on PI-PLCsolubilized GPI-anchored proteins. In vivo, solubilization of p76 is necessary to reveal its proteolytic activity (Braun-Breton and Pereira da Silva 1988). These results suggested the presence of a compartmentalized, developmentally regulated PI-PLC in the parasite. This type of enzyme regulation may be a feature of all plasmodia, as
453
GPI-PHOSPHOLIPASE
we have recently identified in the rodent malaria parasite Plasmodium chabaudi a similar serine protease activity that also depends on cleavage by PI-PLC (BraunBreton et al. submitted for publication). In this paper, we report the characterization of a developmentally regulated, Ca*‘independent, GPI-degrading activity from both P. falciparum and P. chabaudi. This activity resolves into a phospholipase C (PLC) and a phospholipase A (PLA). The GPI-PLC could be responsible for the in vivo induction of the previously described serine protease activity which correlates with merozoite maturation and red blood cell invasion. The biochemical characterization of these enzymes and eventual cloning of the corresponding genes should aid in the clarification of their physiological role. MATERIALS Strains chabaudi
AND
METHODS
and parasite production. Clone F of the P. IP-IPCl strain was injected into 2-month-old BALB/c mice. The strain characteristics and
female the methods used for its culture and synchronization have been previously described (Falanga et al. 1984). Red blood cells (RBC) were recovered from infected mice when the parasitemia was 50% and osmotically lysed as described (Braun-Breton et al. 1986). Synchronous cultivation of the P. falciparum Palto Alto FUP-Uganda strain was performed (Braun-Breton et al. 1988) and trophozoite-infected RBC were purified on a Percoll-sorbitol gradient prior to further cultivation. The maturation of schizonts was followed by Giemsa staining and the concentrated schizonts were lysed by osmotic shock (Braun-Breton et al. 1986). Preparation of protein extracts. Parasite proteins were extracted by incubation of osmotically lysed cells in 1 vol of 2% NP-40, 100 mMTris, pH 8.0, and 10 mM EDTA for 20 min at room temperature. Preparation of [9,10(n)-3H]myristate-labeled mflSG. Trypanosomes of the Botat-1 variant of Trypanosoma equiperdum (Capbem et al. 1977) were pu-
rified from the blood of C3H-infected female mice (Hereld et al. 1988b). [‘H]Myristate labeling of trypanosomes and isolation of the membrane-form variant surface glycoprotein (mfVSG) were performed following the protocol described by Hereld et al. (1988b). The [3H]mfVSG preparation (12 mg/ml) was stored at -80°C. GPI-anchor-degrading
activity assay. GPIdegrading activity was assayed by its ability to cleave
454
BBAUN-BRETON
[3H]mfVSG, generating butanol-extractable 3H materials and unlabeled soluble VSG. Protease-free PIPLC of different origins were tested. These were purified B. cereus PI-PLC (Boehringer) and purified T. brucei GPI-PLC (a kind gift of Dr. P. Overath). NP-40 extracts from P. fulciparumand P. chabaudi-infected red blood cells were assayed. Purified 13H]myristic acid-labeled mfVSG (2 ug = 40 pmol, 7 x lo3 cpmug) was incubated with PI-PLC or parasite extract in 1% NP-40, 50 mM Tris, pH 8.0, 5 mM EDTA, for 30 min at 37°C. The released [3H]DAG was extracted with 500 ul of butanol (water saturated). The radioactivity in 400 pl of the organic phase was determined by liquid scintillation counting. When required the Triton X-l 14 partition of P. chabaudi proteins was performed as described (Bordier 1988). Phosphatidylinositol (PZ)-degrading activity. The activity was assayed by its ability to cleave L-3phosphatidyl-[2-3H]inositol, generating water-soluble 13H]inositol, butanol-extractable product, and uncleaved [3H]PI. [3H]PI (10 uM solution, 10-15 Cii mmol) was diluted 1:300 in 1% NP-40,50 mM Tris, pH 8.0, 5 mM EDTA (60,000 cpm per assay) and incubated for 30 min at 37°C with NP-40 extracts from P. falciparum and P. chubaudi merozoites, together with the different PI-PLC preparations. These were B. cereus PI-PLC (from Boehringer), B. thuringiensis PIPLC (from Immunotech), T. brucei PI-PLC (from Dr. P. Overath), and phospholipase A, (from Sigma). Butanol extraction was then performed and the radioactivity was determined by liquid scintillation counting of the aqueous and organic phases. Thin-layer chromatography. The products released from [3H]mfVSG in the butanol phase (400 ul), by Listeriu monocytogenes PI-PLC and heat-inactivated (2 hr at 60°C) human serum (phospholipase D) taken as controls as well as P. chabaudi merozoite and P. falciparum blood stage extracts were analyzed by thinlayer chromatography. After evaporation of the solvent, the extracted material was redissolved in 40 ul of CHCl&H,OH/H,O (10:10:3) and chromatographed on a thin-layer Silica Gel 60 (Merck; heated 1 hr at 100°C prior to use) using CHCl,/CH,OH/H,O (65:25:4) as solvent. The plate was sprayed with EN3HANCE (New England Nuclear) and fluorographed. Commercially available sn-1-stearoyl-2-[i4C]arachidonylglycerol, L-3-phosphatidyl-[2-‘Hhnositol and [9,10(n)‘Hlmyristic acid (Amersham) were used as controls for migration of DAG, PI, and myristic acid. Phenylboronate column chromatography of GPZdegrading activity. Detergent extracts of P. fulciparum or P. chabuudi proteins were applied to a phe-
nylboronate (Sigma) column (10 vol beads/l vol extract). The bound proteins were eluted by subsequent addition of 10 mM sorbitol, 100 mM sorbitol, 1 M guanidine, and 3 M guanidine. Starting with 100 mg of parasite proteins, about 10 mg bound to the phenylboronate column and 1 mg was recovered in the 10 mM
ETAL. sorbitol eluate, 2 mg in the 100 mM sorbitol eluate, 2 mg in the 1 M guanidine eluate, and 4 mg in the 3 M guanidine eluate. The different eluates were concentrated to the volume of the extract applied to the column and dialyzed for 48 hr at 4°C against 1% NP-40,50 mM Tris, pH 8.0, 5 mM EDTA in the presence of 50 &ml leupeptin and pepstatin. Zmmunoprecipitation activity. Twenty to
of the GPZ-anchor-degrading
40 ul of detergent-extracted parasite proteins was incubated with 10 to 20 ul of Maps (Bio-Rad kit)-purified, protease-free IgG in 50 mM Tris, pH 8.0, 5 mM EDTA, and 1% NP-40 for 1 hr at 37°C. After the addition of 200 ul of 10% IgSorb (The Enzyme Center, Inc., Malden, MA 02148) and incubation for a further 30 min at 4°C the suspension was centrifuged. The supematant was assayed for GPIanchor-degrading activity. Western blot analysis. Western blots of phenylboronate column eluates from P. fulcipurum and P. chabaudi were performed using a monoclonal antibody, mAT3, raised against T. brucei GPI-PLC (Btilow and Overath 1986). The antigens were visualized after incubation with anti-mouse alkaline phosphataselabeled immunoglobulins and reaction with nitro blue tetrazolium substrate (Promega Biotec). RESULTS
To characterize the PI-PLC-like activity present in P. falciparum and P. chabaudi parasites, we used an enzyme assay similar to that described for T. brucei (Hereld et al. 198813). The assay estimates the ability of 1% NP-40 extracts of P. falciparum and P. chabaudi to liberate [3H]myristate-labeled DAG from mfVSG of T. equiperdum. As shown in Table I the GPI-degrading activity was first detected at the end of the intraerythrocytic cycle and increased as young schizonts (less than four nuclei) matured into merozoites. The P. falciparum activity was essentially recovered in the sedimentable fraction and not in the soluble fraction (Table II). This suggests that the activity is membrane associated. The extensive washing of the membrane fraction to avoid contamination by soluble components is probably responsible for the loss of activity during the separation. In phase separation experiments with Triton X-114, the P. chabaudi GPI-degrading activity behaved as if it were membrane bound (Table II). Taken together these results suggest
CHARACTERIZATION
OF PLASMODIA
Time Course of GPI-Anchor-Degrading
TABLE I Activity in P. falciparum
Protein concentration
Source of enzyme VSG lipase
GPI-PHOSPHOLIPASE
Activity (pmoVml/min)
15 p&ml
455
and P. chabaudi
Specific activity (pmol/mg/min)
3750
2.5 x 105
Activity relative to schizonts P-N)
P. falciparum
NP-40 extracts Mature trophozoites Schizonts, ~4 nuclei Schizonts, 4-8 nuclei Schizonts, 38 nuclei Merozoites
5 5 5 5 2
mglml mg/ml mg/ml mg/ml mg/ml
25 150 500 750 600
5 30 100 150 300
0.05 0.3 1 1.5 3
5 5 5 5
mg/ml mg/ml mg/ml mg/ml
100 500 1000 1500
20 100 200 300
0.2 1 2 3
P. chabaudi
NP-40 extracts Schizonts, ~4 nuclei Schizonts, 4-8 nuclei Schizonts, 28 nuclei Merozoites
Note. The activities presented correspond to the average values of at least three independent experiments.
that the P. falciparum and P. chabaudi phospholipases are membrane-bound proteins. Next, using the same assay, we determined the effects of cations, thiol reducing TABLE II Relative GPI-Degrading Activity in Soluble and Membrane Fractions
Fraction
Relative activity (%I
P. falciparum
NP-40 Soluble Membrane associated
100 co.05 40-50
P. chabaudi
NP-40 Aqueous phase Triton phase
100 co.05 SO-90
Note. P. falciparum parasites were lysed by osmotic shock and soluble and membrane fractions were separated by differential centrifugation (Braun-Breton et al. 1986) or by direct extraction of proteins with 1% NP-40. The relative activity of each fraction was compared. For P. chabaudi osmotically lysed parasites, the comparison was made between a Triton X-l 14 partition of soluble and membrane components and a direct extraction by 1% NP-40.
agents, detergents, and protease inhibitors on the P. falciparum and P. chabaudi phospholipases. These effects were compared to those obtained with T. brucei GPI-PLC and B. cereus PI-PLC in the same enzyme assay (Table III). The plasmodia phospholipases are similar to T. brucei GPI-PLC in a number of characteristics. First, Ca*’ and EDTA had no effect on their activity, indicating that they are calcium-independent, GPI-degrading activities. Second, the activities exhibited an acute sensitivity to Zn*+ ions and to pCMPS and a marked stimulation by DTT, characteristics which are consistent with the requirement for a reduced sulfhydryl group for maximal activity. Third, the liberation of [3H]myristatelabeled fatty acid was sensitive to sodium deoxycholate and slightly sensitive to Triton X-100. Finally, the activity of the different phospholipases was unaltered by the presence of the protease inhibitors leupeptin and pepstatin. The T. brucei phospholipase C appears to be specific for GPI as it cleaves efficiently the VSG glycolipid anchor, but cleaves PI
456
BRAUN-BRETON
ET AL.
TABLE III The Effects of Various Reagents on the GPI-Anchor-Degrading
Activities of Different Phospholipases
% GPI-degrading P. falciparum
P. chabaudi
activity T. brucei
B. cereus
None
100
loo
100
100
CaCI, 2mM 10 mM
100 80
100 90
100 95
nt 90
80
nt
100
loo
85 95
nt 100
EDTA 2mM 10 mM
pCMPS 5mM
I
10
20
100
ZnCI, 2mM IO mM
50 30
nt 2
90 0
100 2
DTT 2mM 10 mM
140 200
130 150
170 205
loo 100
Triton X-100 0.2% 1%
100 66
nt 40
95 70
100 70
15 2
25 2
15 2
85 70
100
100
100
100
Nadeoxycholate 0.2% 1% Leupeptin pepstatin
P. falciparum and P. chabaudi GPI-degrading activities were measured in 1% NP-40 extracts. The T. GPI-PLC (2.5 x lo-* U/ml) and B. cereus PI-PLC (2.5 x lo-* U/ml) were purified enzymes. The numbers represent the activity as a percentage of controls and correspond to the average value of three independent experiments. Note. brucei
to a much lesser extent (Btilow and Overath 1986; Hereld et al. 1986; Fox et al. 1986). The P. fulciparum extract contains a phospholipase C activity, since the p76 serine protease exhibits a CRD epitope after its cleavage by this enzyme (Braun-Breton et al. 1990). The products of [3H]mfVSG degradation by a P. chubaudi merozoites extract and a P. fulcipurum extract were analyzed by thin-layer chromatography. Both plasmodial enzymes and the L. monocytogenes PI-PLC (Mengaud et al. 1991) generate products which comigrate with control diacylglycerol (Fig. 1). These results confirm the presence of a GPI-PLC in P. fulciput-urn and P. chubuudi extracts. A second
and major product, comigrating with myristic acid, is generated by the plasmodial extracts. The enzyme(s) responsible for the generation of myristic acid has not been identified but should be a PLA activity. Depletion of the plasmodial GPI-degrading activity was obtained with different inhibitors active on various PI-PLCs but not reported to be active on known PLAs; depletion of the activity was also obtained by immunoprecipitation with antibodies to T. brucei GPI-PLC (Table VI). This suggests that the generation of myristic acid might be the consequence of further degradation (by the PLA activity) of the diacylglycerol generated by the parasite PLC. We cannot ex-
CHARACTERIZATION
OF
PLASMODIA
DAG
myrtsttc
acid
+
3
origin
j
1. Products of the P. chabaudi and P. falciGPI-anchor-degrading activity. Thin-layer chromatography of butanol-extracted material released from [3H]mfVSG was performed on Silica Gel 60 plate as described (under Materials and Methods). Lane 1, sn-1-stearoyl-2-[r4C]arachidonylglycerol used as a control for migration of DAG; lane 2, L-3phosphatidyl-[2-3H]inositol; lane 3, [9,10(n)-3H]myristic acid; lane 4, material released by a P. chabaudi merozoites extract; lane 5, material released by a P. fulciparum blood stage extract; lane 6, material released by human serum phospholipase D (2 hr at 60°C heat-inactivated human serum); lane 7, material released by control L. monocytogenes LO28 culture supematant; lane 8, same as lane 7 with a deposit twice that in lane 7.
FIG. parum
elude, however, that the parasite PLA activity is Ca2+ -independent, membrane associated, sensitive to thiol reagents and detergents, and immunoprecipitated by anSerum tibodies to T. brucei GPI-PLC. phospholipase D produces a different species, presumably phosphatidic acid, migrating close to the loading point of the samples (Davitz et al. 1987). In some extracts, contamination with phospholipase D (probably from host serum) was detected (data not shown).
451
GPI-PHOSPHOLIPASE
We have also investigated the possibility that the plasmodial phospholipases, like the T. brucei enzyme, are GPI specific. L3H]mfVSG and phosphatidyl-[3H]inositol were compared as substrates for different phospholipases. Phospholipase C digestion of both substrates would result in the release of radioactive material which would partition differently from the substrate: [3H]inositolphosphate into the aqueous phase, versus phosphatidyl-[3H]inositol into the organic phase and likewise [3H]DAG into the organic phase, versus [3H]mfVSG into the aqueous phase. As shown in Table IV, bacterial PI-PLC are able to digest both substrates. In contrast, PI is a poor substrate for both the trypanosomal and malarial enzymes. Phospholipase A2 was used as a control in these experiments. Phospholipase A2 digestion of phosphatidyl[3H]inositol yields myristic acid and monoacylated phosphatidyl-[3H]inositol, both partitioning in the organic phase like the uncleaved PI. As expected, no release TABLE
IV
f3HlPI cpm Substrate phospholipase
released in the aqueous phase
[3H1mfVSG cpm released in the organic phase
2300-4020
28X1-7000
5500-5960
4OOOA500
