Biochimica et Biophysica Acta, 1091 (1991) 15-21 © 1991 Elsevier Science Publishers B.V. (Biomedical Division) 0167-4889/91/$03.50 ADONIS 0167488991000682
15
BBAMCR 12824
Release of a membrane surface glycoprotein from human platelets by phosphatidylinositol specific phospholipase(s) C Animesh Dhar and Shivendra D. Shukla Department of Pharmacology, Unioersity of Missouri, School of Medicine, Columbia, MO (U.S.A.) (Received 6 August 1990)
Key words: Phosphatidylinositol specific phospholipase C; Glycoprotein: GPlb; Platelet membrane: Phospholipase C; (Human)
Phosphatidylinositol (PI) specific phospholipase C (Plase C) treatment of human platelets caused release of a surface glycoprotein in the medium. Human blood platelets tvere isolated by low speed centrifugation and surface glycoproteins were labelled with periodate/[aHiborohydride procedure. Intact surface-labelled platelets were treated with P|ase C purified from culture filtrates of Staphylococcus aureus (SA) or Bacillus thuringiensis (BT). After Plase C treatments platelets were spun at low speed, pellet and supernatant were separated. The supernatant was further centrifuged at high speed (140000 × g) for 30 rain. The resulting supernatant and the pellet from low speed were subjected to SDS-PAGE analysis. Protein patterns were obtained by fluorography. Release of a specific glycoprotein of approx. 150 kDa in the medium was observed due to the Plase C treatment. Prolonged incubation of platelets in 0.25 M sucrose and depletion of NaC! concentrations also affected the release of this glycoprotein. BT-Plase C released more approx. 150 kDa protein than SA-Plase C. Western blot experiment with a monocional antibody (mAB), epitope SZ2, reactive to human platelet surface glycoprotein ib (GPib) complex, confirmed that released 150 kDa glycoprotein reacted with mAB of GPlb. The release of this protein by Plase C was not inhibited by proteinase inhibitors (EDTA, PMSF and leupeptin). Treatment of human platelet membranes with Plase C also caused release of this glycoprotein as evidenced by reactivity to GPIb-mAB. These studies demonstrate that Plase C treatment causes release of 150 kDa glycoprotein from human platelet membrane surface. It is suggested that 150 kDa glycoprotein is anchored to P| in human platelets and that this glycoprotein represents the GPlb complex.
Introduction
Exposure of various cells and tissues to phosphatidylinositol specific phospholipase C (PIase C) has been shown to cause selective release of membrane proteins in free and soluble form. This led to the suggestion that these enzymes are specifically linked to PI in membranes [1,2]. Recent developments have established that a number of cell surface proteins are attached to the membrane by covalent linkage to a glycosyl-phosphatidylinositol anchor [2-5]. The physiological role of this association is less known; although it is suggested that it might serve novel functions in addition to retaining proteins at the cell surface [6-12]. It is worth noting that PIase C is widely distributed in mammalian tissues [1] and that such enzymes are directly involved in the receptor-coupled generation of second messengers via phosphoinositide turnover [13]. Correspondence: S.D. Shukla, Dept. Pharmacology, University of Missouri, School of Medicine, Columbia, MO 65212, U.S.A.
Human blood platelets play an important role in thrombosis and hemostasis. Platelet surface glycoproteins play roles as being receptors for thrombin [14], von Willebrand factor [15] and fibrinogen [16] and are also functional components of the haemostatic process, e.g., in platelet-endothelium interaction. We report here the results of a systematic study of the release of a glycoprotein (approx. 150 kDa) from platelet surface by PIase C. Part of this work was reported in an abstract of The Joint Meeting American Society for Biochemistry and Molecular Biology and American Society for Cell Biology held January 29-February 2, 1989, San Francisco, California (see Ref. 17). Materials and Methods
Materials NaB3H4 was bought from ICN Radiochemicals, CA. PAF (1-O.hexadecyl- 2-sn-glycerol- 3-phosphorylcholine ) was supplied by Bachem (Torrance, CA) and was routinely checked for its purity by silica gel thin-layer
16 chromatography in a solvent system containing chloroform/ methanol/water (65 : 35 : 6, v/v). Thrombin was supplied by Sigma Chemical (St. Louis, MO). Triton X-114 was bought from Calbiochem (San Diego, CA). Monoclonal antibody of human platelet glycoprotein lb, epitope SZ2, was supplied by Amac (Westbrook, ME). All other chemicals and solvents used were of the analytical reagent grade.
Isolation of human platelets Blood (50-60 ml) from healthy donors, who had not taken any medication for at least 2 weeks, was collected in vacutainer tubes containing acid-citrate-dextrose (Beckton Dickinson) and was immediately centrifuged at 200 × g for 8 rain at 24°C to obtain platelet-rich plasma (PRP). PRP was centrifuged at 750 × g for 10 rain and the platelet pellet was suspended in either phosphate (10 mM)-buffered saline (155 mM) medium (PBS) containing 2 mM EDTA (pH 7.5) or 0.25 M sucrose containing 2 mM EDTA (pH 7.5), as appropriate. Purification of phosphatidylinositol specific phospholipase C (Plase C) PIase C was purified from culture supernatants of Staphylococcus aureus [SA] [18,19] or Bacillus thuringiensis [BT] in our laboratory using Q-sepharose (Pharmacia Chemical) followed by Sephadex G-75 (Pharmacia Chemical) and HPLC (Mono-Q) chromatography. The purity of the PIase C was checked by SDS-PAGE [20] and showed a single band corresponding to approx. 33 kDa (unpublished data). Surface labelling of platelets with periodate/[3H]boro hydride and their treatments with Plase C Washed platelets suspended in PBS containing 2 mM EDTA were incubated with 2 mM sodium metaperiodate (Sigma Chemical) for 10 rain in the cold (4°C). Platelets were subsequently washed thoroughly with PBS-EDTA buffer by centrifugation and were incubated for 45 rain with 0.5 mCi of NaB3H4 in PBSEDTA buffer [21,22]. Cells were again washed and resuspended in PBS-EDTA or 0.25 M sucrose-EDTA buffer for treatments with BT- or SA-PIase C, respectively. Intact surface-labelled platelets were treated with BT- or SA-Plase C at 37°C for desired time periods and were briefly centrifuged (12 000 × g for 1 min), the resultant supernatant and pellets were separated. The pellet was retained for further use. The supernatant was recentrifuged at 140000 × g for 30 rain. Supernatant was dialyzed exhaustively to eliminate salt or sucrose as necessary and was concentrated by blowing N 2 at 50°C. Pellets and the concentrated supernatant were solubilized in a buffer containing 3~ sodium dodecylsulphate (SDS), 5~ 2-mercaptoethanol, 10~ glycerol, 0.0625 M Tris at pH 6.8 [20]. The samples were subjected to
SDS-polyacrylamide slab gel electrophoresis using 7.5% acrylamide under the conditions described by Laemmli [20]. The gel was stained with Coomassie brilliant blue, destained and soaked in Amplify (Amersham Corporation, IL) for 30 rain with agitation. The gel was dried under vacuum at 80°C using a Bio-Rad slab gel dryer (Model 224, Richmond, CA), exposed to Kodak-X-Omat film at - 7 0 ° C and the fluorograph developed after 48 h exposure. The regions of gel corresponding to radioactive bands were cut out and dissolved in 0.1 ml of 60~ (w/v) perchloric acid and 0.2 ml of 30~ (v/v) H202 for 2 h at 90°C [23]. The dissolved material was then allowed to cool and their radioactivity determined using 10 ml of scintillation cocktail.
Phase separation in Triton X-I I 4 solution After surface labelling platelets with [3H]borohydride, 1~ Triton X-114 [24] Tris-buffered saline (TBS) (pH 7.4), was added in the medium containing platelets (1. 109/ml). It was mixed thoroughly and kept on ice for 30-40 min. The suspension was spun at 4°C and the supernatant transferred to a 1.5 ml conical Eppendorf microfuge tube. This tube wa~ then incubated for 5 rain at 37 ° C. The clear solution became cloudy. It was centrifuged at 500 × g for 3 rain and the detergent appeared as an oily droplet in the bottom, while the upper phase contained the aqueous fraction. The upper aqueous fraction was collected, 0.9 ml of TBS was added in the lower phase, mixed well and kept at 4 °C (ice) for 30 rain. It was again phase-separated by incubation at 37°C followed by centrifugation. The upper phase and the lower detergent phases were collected and concentrated under N2 at 50°C. These samples were solubilized in SDS-cocktail, subjected to SDS-PAGE for separation of proteins and processed for fluorography. After 48-72 h exposure on Kodak X-ray film it was developed to monitor protein patterns. Electrophoretic immunoblotting method After SDS-PAGE of samples, proteins were electrophoreticaUy transferred onto nitrocellulose sheets (BioRad, California) [25]. The electrode buffer contained 25 mM Tris, 192 mM glycine, methanol 20~ (v/v) at pH 8.3. The transfer to nitrocellulose paper was done at 100 V (360 mA) for 5-6 h in cooling conditions. The electrophoretic blots were soaked with 3~ gelatin in Tris-saline (0.9~ NaCI/10 mM Tris-HCl (pH 7.4)) buffer (TBS) for 1 h at room temperature to saturate additional binding sites. The sheets were washed with TBS containing 0.05~ Tween 20 (pH 7.4) and were incubated with monoclonal antibody (mAB) of human platelet glycoprotein lb (GPIb), epitope SZ2 [26], followed by incubation with goat antimouse IgG horseradish peroxidase conjugate (Bio-Rad). The color of the blot was developed using HRP color developing reagent (Bio-Rad) and a photograph was taken immediately.
17
Isolation of platelet membranes Human platelets were separated as described earlier and were suspended in phosphate (10 mM)-buffered saline (155 raM) containing 2 mM EDTA, 100 #M PMSF and 20 #M leupeptin. Platelets were sonicated using a Vibro cell sonicator (Sonics and Materials, Danbury, CT) at setting 3 for eight cycles at 15 s intervals at 4 ° C. The suspension was then centrifuged at 4°C at 600 × g for 10 min. The resultant supernatant was further centrifuged at 140000 × g for 1 h at 4°C. The pellet was resuspended in the PBS buffer containing proteinase inhibitors (EDTA, leupeptine and PMSF), was treated with BT-PIase C at 37°C and processed by procedures described earlier. The samples were subjected to SDS-PAGE and Western blot analysis with mAB of GPIb as described above. Results and Discussion
Fluorography of surface-labelled platelets treated with SA-Plase C or BT-Plase C Human platelets were treated with SA-PIase C for various time periods (0, 15 and 30 rain). The supernatant and pellet fractions (see Materials and Methods) were analyzed by fluorography. Examination of the pattern showed that there is a significant release of a glycoprotein of Mr approx. 150 kDa in a time- and dose-dependent manner (data not shown). The release of the protein by SA-PIase C (5 U / m l ) was maximum during a 30 min time period (Table I). Interestingly, the release of this protein was also observed in the control (i.e., no PIase C) which contained 0.25 M sucrose medium. This background release was less than that after PIase C treatment. It was necessary to use a sucrose medium since SA-PIase C is inhibited by salts and ions [18]. On the basis of this observation an increasing concentration of NaCI was added to a de-
creasing concentration of sucrose to find out whether the release of this protein was due to the depletion of NaCl in the medium. Incubation of surface-labelled platelets with NaC1 for 30 rain did not release this protein significantly but a gradual increase in the concentration of sucrose in the medium with a concomitant decrease in NaCl caused significant release of the 150 kDa glycoprotein. Sucrose (0.25 M) alone released a significant amount of this glycoprotein in 30 min which was about 15% over 0 min control (results not shown). It was also observed that 0.25 M sucrose EDTA buffer and SA-PIase C did not alter the thrombin-stimulated aggregation response [7] or [3H]serotonin secretion (results not shown). These results indicate that depletion of sodium in the medium caused release of a glycoprotein which was also released by SA-PIase C. Therefore, we next used BT-PIase C which is known to be less sensitive to inhibition by sodium chloride than SA-PIase C [27]. Surface-labelled platelets were treated with BT-PIase C (0.5 U/ml) in PBS-EDTA and the release of proteins was monitored using the protocol described in the Methods. In this case also the same 150 kDa glycoprotein was released into the medium after treatment with BT-PIase C (Fig. 1). The release of this glycoprotein by BT-PIase C was compared with SAPIase C and observed that BT-PIase C released a highly significant amount of 150 kDa glycoprotein (P < 0.001) in the medium, whereas SA-PIase C released 10-15% ove~ the control (Table I). BT-PIase C also released this protein more at a concentration of 0.5 U / m l (23%) than 0.25 U (16.0%) or 0.10 U (13.0%) which suggested that the release of this protein from platelet surface by BT-PIase was in a dose-dependent manner (data not shown). The other areas of the gel were cut and counted for radioactivity and no significant release of other components was observed by PIase C treatment (data not shown). This indicates that PIase C treatment re-
TABLE I
Release of glycoprotein by Plase C isolated from Staphylococcus aureus (SA) or Bacillus thuringiensis (BT) Washed human platelets were surface-labelled by sodium borohydride procedure (see Materials and Methods), then treated with BT or SA-PIase C at 37 o C for 30 min. Control platelets were treated with PBS-EDTA buffer (pH 7.2). The platelets were centrifuged and supematant and pellet were separated (see Materials and Methods for details) and were further subjected to SDS-PAGE and fluorography. The regions of gel corresponding to radioactive bands were cut out, dissolved in perchloric acid and H202 and radioactivity counted in 10 ml scintillation cocktail (see Materials and Methods). Values are expressed as the mean_+ S.E.M. for three separate experiments. Figures in the parentheses indicate percentages. Radioactivity (cpm) Control
SA-Plase C (5.00 U/ml)
BT-Plase C (0.50 U/ml)
0 min
30 rain
0 min
30 min
0 rain
30 min
Pellet
28642_+3187 (93.9)
28822_+3240 (91.3)
356234-5733 (91.2)
27901+2958 (84.0)
30508_+3254 (90.3)
27326_+3569 (74.4)
Supernatant
1846+ 158 (6.1)
3720+ 432 (8.6) [A]
34144- 479 (8.7)
5311_+ 427 (15.9) in]
3249_+ 298 (9.6)
Significance: A vs. B, P < 9.05, A vs. C, P < 0.001.
8 433 +
(23.6)
It]
572
18
significantly in the aqueous phase after treatment of surface-labelled platelets with BT-PIase C for 30 rain with a concomitant decrease of radioactivity in the detergent phase. Although the radioactivity in the aqueous phase of control was high (61~), the aqueous phase from PIase C-treated platelets showed radioactivity (87%) which was higher than control. Newman et al. [28] and Clemetson et al. [29] observed that one of the platelet surface glycoprotein, GPIb (150 kDa) partitioned in aqueous phase in Triton X-114 method. Our results of high level of 150 kDa radioactivity in the Triton X-114 aqueous phase in control confirm their observations. Physiological stimuli like thrombin (1 U/ml) and PAF (10 -7 M) slightly increased the radioactivity in the aqueous phase with a concomitant decrease in radioactivity in the detergent phase in comparison to control platelets (Table If). Phase separation in the Triton X-114 solution experiment demonstrated that a high amount of 150 kDa protein partitioned in the aqueous phase without any treatment (i.e., in control). It was earlier observed by Newman et al. [28] and Clemetson et al. [29] that GPIb is found in the Triton X-114-aqueous phase and this may be due to the very high sugar content of GPlb [30,31]. Although the additional appearance of radioactive proteins in aqueous fractions (26% over control) was observed after PIase C treatment, this method has limitations due to partition!ng of most of the 150 kDa protein in the aqueous phase. Thrombin and PAF treatment also showed a slight increase in 150 kDa [aH]glycoprotein in the aqueous phase and these might represent protein released after PI cleavage caused by activation of endogenous Plase C by these stimuli.
150K
Fig, 1, Release of [JHlglycoprotein by BT-Plase C treatment of surface-labelled human platelets. [3H]Borohydride-labelled platelets were treated with BT-Plase C (0.5 U/ml) for 30 min at 37°C and pellets and supernatants were subjected to SDS-PAGE followed by fluorography (see Materials and Methods). (A) 0 min control pellet: (B) 0 min control supernatant; (C) 30 min control pellet: (D) 30 min control supematant; (E) 0 min Plase-treated pellet: (F) 0 min PlaseC-treated supernatant; (O) 30 min Plase C-treated pellet: and (H) 30 rain Plase C-treated supernatant.
leased specifically 150 kDa protein from platelet surface under the experimental conditions described above.
Phase separation with Triton X-I I4 A solution of nonionic detergent Triton X-l14 is homogeneous at 0°C but it separates in an aqueous phase and detergent phase at 20°C or above. This has provided a novel method-for partitioning hydrophilic and hydrophobic proteins [24]. Table II demonstrated that radioactivity of 150 kDa protein was increased
Western blot analysis probed by a monoclonal antibody of GPIb, epitope SZ2 Platelet membrane glycoprotein GPIb is most prominently labelled by the [3H]borohydride procedure used in this study and it also migrates in the same region
TABLE !1 Phase separation in Triton X-114 solution
Surface glycoprotein-labelled human platelets were treated with 1% Triton X-l14 and the detergent and aqueous phases were separated (see Materials and Methods). The other conditions of this experiment were same as in Table 1. In some experiments before, Triton X-114 phase separation platelets were treated with thrombin (IU) or PAF (10 -7) for 1 min. Values are expressed as the mean 4- S.E.M. from two separate experiments. Figures in the parentheses indicate percentages. Radioactivity (cpm) of approx. 150 kDa protein 30 rain
1 rain
control
BT-Plase C
control
thrombin (1 U)
PAF (10- 7)
Aqueous
5 822 4- 632 (60.9) [A]
9133 4-1045 (87.0) [B]
5 969 4- 617 (73.1) [CI
6 855 4- 599 (82.9) [DI
7 752 4- 666 (81.8) [E]
Detergent
3 723 + 423 (39.0)
1364 + 126 (12.9)
2191 + 238 (26.8)
1406 + 135 (17.0)
1715 + 194 (18.1)
Significance: A vs. B, P < 0.05: C vs. D, P < 0.05; C vs. E, P < 0.05.
19 (approx. 150 kDa) [21,22]. We, therefore, examined the reactivity of GPIb mAB to the released surface glycoprotein. Platelets were treated with BT-PIase C (0.5 U,/ml) for 30 min and then centrifuged (see Materials and Methods). The pellet and supernatant of platelets treated with PIase C or without PIase C (control) were solubilized and separated by SDS-polyacrylamide gel electrophoresis. Polypeptides were transferred from gel to nitrocellulose sheets and the sheets were treated with 1 : 1000 dilution of mAB human platelet GPIb, epitope SZ2. The color of the blot was developed by incubation with goat antimouse horseradish peroxidase followed by color reaction (see Materials and Methods). The photograph was taken immediately. Examination of the blot (data not shown) showed that PIase C-treated supernatant reacted strongly with mAB of GPIb in comparison to control. Control supernatant showed little reactivity with GPIb mAB, epitope SZ2. This method was repeated without mAB of GPIb and the reaction of nitrocellulose sheet was negative (results not shown) which ruled out the nonspecific reaction of the released protein. Similar results (Fig. 2) were also observed when the supernatant was obtained from platelets treated with BT-PIase C (0.5 U/ml) in a buffer medium containing proteinase inhibitors EDTA (2 mM), phenyl methyl sulfonyl fluoride [PMSF] (100/~M) and leupeptin (20 /~M). These agents inhibit a wide variety of proteases, including calcium or metal-dependent proteinases (e.g., calpain) and serine and/or thiol proteinases. The positive reactivity with the monoclonal antibody of GPIb, epitope SZ2 suggested that released protein by PIase C treatment represented GPIb complex of platelet surface glycoprotein. It was also established that the glycoprotein released by PIase C action on platelets was not due to the activation of proteinases and that the PIase C activity is unaffected by proteinase inhibitors. Treatment of human platelet membranes with BT-Plase C We next investigated whether the released protein from platelet surface by PIase C is a membrane component or cytosolic contaminant or any proteolytic product of platelet protein. We therefore isolated platelet membranes and treated with BT-PIase C in the presence of proteinase inhibitors (see Methods) and the released surface glycoproteins were examined by immunoreactivity of GPIb mAB, epitope SZ2. It was observed that PIase C-treated membrane fraction supernatant was highly positive, whereas control membrane fraction supernatant showed no reactivity with monoclonal antibody to GPIb (Fig. 3). This observation suggested that approx. 150 kDa component is directly released from the membrane surface of human platelets due to PIase C treatment. In experiments where platelets membranes were used, control supematant showed a band w~ich had faint reactivity to mAB to GPIb and
A
B
C
D
150K
Fig. 2. Western blot analysis of BT-PIase C-treated human platelets probed by a monoclonal antibody of GPIb, epitope SZ2. BT-PIase C (0.5 U / m l ) treatment of intact human platelets was conducted in a medium containing proteinase inhibitors, i.e., EDTA, leupeptin and PMSF (see Results). The samples were then subjected to SDS-PAGE and were electrophoretically transferred to a nitrocellulose sheet. The sheet was then treated with mAb of human platelets GPIb, epitope SZ2, followed by incubation with gout antimouse IgG horseradish peroxidase conjugate. The color of the blot was developed and a photograph was taken immediately. (A) Control pellet; (B) Control supernatant; (C) BT-PIase C-treated pellet; (D) BT-Plase C-treated supernatant.
migrated below the level of approx. 150 kDa (Fig. 3). This band was not seen reproducibly and therefore is not significant. The released 150 kDa glycoprotein reacted with a mAB of human platelet surface GPIb and confirms that the released glycoprotein represents GPIb complex of platelet surface. GPIb has been postulated to be receptor for thrombin in human platelets [14J and also to be involved in vWF-mediated adhesion of platelets to the subendothelium [15]. The epitope SZ2 was directed against GPIb complex and inhibited ristocetin induced platelet agglutination and also collagen- and PAF-induced platelet aggregation [26]. At present, physiological significance of this released 150 kDa protein is unknown and remains to be investigated. It has been suggested that platelet storage results in a redistribution of GPIb molecules and this has provided evidence for a large intraplatelet pool of GPIb [32]. However, our study indicated that the source of released protein is not from the internal pool of GPIb because PIase C caused release of this glycoprotein
20 A
B
C
D
15OK
the 150 kDa protein might reside in the membrane in two types of interactions. A fraction of this protein (e.g., 15-20~) might be exclusively attached to the PI, while the remainder might be putative transmembrane in nature or is somehow resistant to Plase C action. If so, the physiological significance of the differential interactions of the same protein with PI in membranes will be an interesting issue to resolve.
I
Acknowledgements We are grateful to cal assistance and to manuscript. S.D.S. is Development Award
Mr. Lance Antle for expert techniMs. Kristin Nelson for typing the the recipient of a Research Career from NIH (DK 01782).
References 1 2 3 4 5
Fig. 3. BT-Plase C treatment of human platelet membranes. Isolated human platelet membranes were treated with BT-Plase C and the pellet and supernatant fractions were obtained (see Materials and Methods). Both fractions were subjected to Western blot analysis as described in Fig. 2. (A) BT-Plase C-treated membrane pellet; (B) BT-Plase C-treated membrane supernatant; (C) Control membrane pellet: (D) Control membrane supernatant.
from isolated membranes of human platelets which are devoid of this pool. It also confirms that the 150 kDa glycoprotein is directly released from the membrane and that it is not a result of cytosolic proteolytic activity or contaminant due to cellular damage after treatment with Plase C, It is important to note that Plase(s) C is highly purified (see Materials and Methods) and has no effect on the release of [3H]serotonin (data not shown) or aggregation of platelets [7]. The effect of Plase C on the release of 150 kDa protein from isolated membrane in the presence of proteinase inhibitors also confirmed that Plase C actions are not the result of proteinase activity. GPlb is a transmembrane protein and how such a protein can be released, by hydrolysis of Pl from outside, remains to be explained. However, it should be pointed out that several other membrane proteins which are putatively transmembrane in nature (e.g., 5' nucleotidase, decay accelerating factor etc.), and interact with cytoskeleton (e.g., decay accelerating factor, N-CAM etc.), are released by Plase C [2,5]. Roberts et al. [33,34] showed that human erythrocyte acetyl cholinesterase anchored to Pl is insensitive to Plase C due to substitution on the inositol ring by plasmanylinositol (palmitoylated on the inositol ring). It may be envisaged that
Shukla, S.D. (1982) Life Sci. 30, 1323-1335. Low, M.G. (1987) Biochem. J. 244, 1-13. Low, M.G. (1989) FASEB J. 3, 1600-1608. Low, M.G. (1989) Biochim. Biophys. Acta 988, 427-454. Ferguson, M.A.J. and Williams, A.F. (1988) Annu. Rev. Biochem. 57, 285-320. 6 Low, M.G. and Saltiel, A.R. (1988) Science 239, 268-275. 7 Shukla, S.D. (1986) Life Sci. 38, 751-755, 8 Selvaraj, P., Rosse, W.F., Silbert, R. and Springer, T.A. (1988) Nature 333, 565-567. 9 Simmons, D. and Seed, B. (1988) Nature 333, 568-570. 10 Huizinga, T.W.J., Van der Schoot, C.E., Jost, C., Klaassen, R., Kleitjer, M., Von dem Borne, A.E.G.K., Roos, D.A. and Tettroo, P.A.T. (1988) Nature 333, 667-669. 11 Low, M.D and Zilversmit, D.B. (1980) Biochemistry 19, 39133918. 12 Mazumdar, R. and Balasubramariam, A.S. (1985) Biochem. Pharmacol. 34, 4109-4115. 13 Abdel-Latif, A.A. (1986) Pharmacol. Rev, 38, 246-272. 14 Okumura, T., Hasitz, M. and Jameison, G.A. (1978) J. Biol. Chem. 253, 3435-3443, 15 Coiler, B,S., Perrschke, E.I,, Scudder, L.E. and Sullivan, C.A. (1983) Blood 61, 99-110. 16 Philips, D.R. and Baughan, A.K. (1983) J. Biol. Chem. 258, 10240-10246. 17 Dhar, A. and Shukla, S.D. (1988) J. Cell. Biol. 107, 55%. 18 Low, M.D. (1981) Methods Enzymol. 71,741-746. 19 Shukla, S,D., Coleman, R. Finean, J.B. and Michell, R.H. (1980) Biochem. J. 187, 277-280. 20 Lameili, U.K. (1970) Nature 227, 680-685. 21 Dhar, A. and Ganguly, P. (1988) Br. J. Haematoi. 70, 71-75. 22 Steiner, B., Clemetson, K.J. and Luscher, E.F. (1983) Thromb. Res. 29, 43-52. 23 BiUoc, F., Heilmann, E., Combrie, R., Boisseau, M.R. and Nurden, A.T. (1987) Biochim. Biophys. Acta 925, 218-225. 24 Bordier, C. (1981) J. Biol. Chem. 256, 1604-1607. 25 Towbin, H., Staehelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354. 26 Ruan, C., Du, X., Xi, X., Castaldi, P.A. and Berndt, M.C. (1987) Blood 69, 570-577. 27 Taguchi, R., Asahi, Y, and Ikezawa, H. (1980) Biochim. Biophys. Acta 619, 48-57. 28 Newman, P.J., Knipp, M.A. and Kahn, R.A. (1982) Thromb. Res. 27, 221-224.
21 29 Clemetson, K.J., Bienz, D., Zahino, M.-L. and Luscher, E.F. (1984) Biochim. B:ophys. Acta 778, 463-469. 30 Okamura, T., Lo~nbart, C and Jamieson, G.A. (1976) J. Biol. Chem. 251, 5950-5955. 31 Judson, P.A., Anstee, D.J. and Clamp, J.R. (1982) Biochem. J. 205, 81-90.
32 Michelson, A.D., Adelman, B., Barnard, M.R., Carroll, E. and Hardin, R.I. (1988) J. Ciin. Invest. 81, 1734-1740. 33 Roberts, W.L., Myher, J.J., Kuksis, A., Low, M.G. and Rosenberry, T.L. (1988) J. Biol. Chem. 263, 18766-18775. 34 Roberts, W.L., Santika,'n, S., Reinhold, V.N. and Rosenberry, T.L. (1988) J. Biol. Chem. 263, 18776-18784.
15
BBAMCR 12824
Release of a membrane surface glycoprotein from human platelets by phosphatidylinositol specific phospholipase(s) C Animesh Dhar and Shivendra D. Shukla Department of Pharmacology, Unioersity of Missouri, School of Medicine, Columbia, MO (U.S.A.) (Received 6 August 1990)
Key words: Phosphatidylinositol specific phospholipase C; Glycoprotein: GPlb; Platelet membrane: Phospholipase C; (Human)
Phosphatidylinositol (PI) specific phospholipase C (Plase C) treatment of human platelets caused release of a surface glycoprotein in the medium. Human blood platelets tvere isolated by low speed centrifugation and surface glycoproteins were labelled with periodate/[aHiborohydride procedure. Intact surface-labelled platelets were treated with P|ase C purified from culture filtrates of Staphylococcus aureus (SA) or Bacillus thuringiensis (BT). After Plase C treatments platelets were spun at low speed, pellet and supernatant were separated. The supernatant was further centrifuged at high speed (140000 × g) for 30 rain. The resulting supernatant and the pellet from low speed were subjected to SDS-PAGE analysis. Protein patterns were obtained by fluorography. Release of a specific glycoprotein of approx. 150 kDa in the medium was observed due to the Plase C treatment. Prolonged incubation of platelets in 0.25 M sucrose and depletion of NaC! concentrations also affected the release of this glycoprotein. BT-Plase C released more approx. 150 kDa protein than SA-Plase C. Western blot experiment with a monocional antibody (mAB), epitope SZ2, reactive to human platelet surface glycoprotein ib (GPib) complex, confirmed that released 150 kDa glycoprotein reacted with mAB of GPlb. The release of this protein by Plase C was not inhibited by proteinase inhibitors (EDTA, PMSF and leupeptin). Treatment of human platelet membranes with Plase C also caused release of this glycoprotein as evidenced by reactivity to GPIb-mAB. These studies demonstrate that Plase C treatment causes release of 150 kDa glycoprotein from human platelet membrane surface. It is suggested that 150 kDa glycoprotein is anchored to P| in human platelets and that this glycoprotein represents the GPlb complex.
Introduction
Exposure of various cells and tissues to phosphatidylinositol specific phospholipase C (PIase C) has been shown to cause selective release of membrane proteins in free and soluble form. This led to the suggestion that these enzymes are specifically linked to PI in membranes [1,2]. Recent developments have established that a number of cell surface proteins are attached to the membrane by covalent linkage to a glycosyl-phosphatidylinositol anchor [2-5]. The physiological role of this association is less known; although it is suggested that it might serve novel functions in addition to retaining proteins at the cell surface [6-12]. It is worth noting that PIase C is widely distributed in mammalian tissues [1] and that such enzymes are directly involved in the receptor-coupled generation of second messengers via phosphoinositide turnover [13]. Correspondence: S.D. Shukla, Dept. Pharmacology, University of Missouri, School of Medicine, Columbia, MO 65212, U.S.A.
Human blood platelets play an important role in thrombosis and hemostasis. Platelet surface glycoproteins play roles as being receptors for thrombin [14], von Willebrand factor [15] and fibrinogen [16] and are also functional components of the haemostatic process, e.g., in platelet-endothelium interaction. We report here the results of a systematic study of the release of a glycoprotein (approx. 150 kDa) from platelet surface by PIase C. Part of this work was reported in an abstract of The Joint Meeting American Society for Biochemistry and Molecular Biology and American Society for Cell Biology held January 29-February 2, 1989, San Francisco, California (see Ref. 17). Materials and Methods
Materials NaB3H4 was bought from ICN Radiochemicals, CA. PAF (1-O.hexadecyl- 2-sn-glycerol- 3-phosphorylcholine ) was supplied by Bachem (Torrance, CA) and was routinely checked for its purity by silica gel thin-layer
16 chromatography in a solvent system containing chloroform/ methanol/water (65 : 35 : 6, v/v). Thrombin was supplied by Sigma Chemical (St. Louis, MO). Triton X-114 was bought from Calbiochem (San Diego, CA). Monoclonal antibody of human platelet glycoprotein lb, epitope SZ2, was supplied by Amac (Westbrook, ME). All other chemicals and solvents used were of the analytical reagent grade.
Isolation of human platelets Blood (50-60 ml) from healthy donors, who had not taken any medication for at least 2 weeks, was collected in vacutainer tubes containing acid-citrate-dextrose (Beckton Dickinson) and was immediately centrifuged at 200 × g for 8 rain at 24°C to obtain platelet-rich plasma (PRP). PRP was centrifuged at 750 × g for 10 rain and the platelet pellet was suspended in either phosphate (10 mM)-buffered saline (155 mM) medium (PBS) containing 2 mM EDTA (pH 7.5) or 0.25 M sucrose containing 2 mM EDTA (pH 7.5), as appropriate. Purification of phosphatidylinositol specific phospholipase C (Plase C) PIase C was purified from culture supernatants of Staphylococcus aureus [SA] [18,19] or Bacillus thuringiensis [BT] in our laboratory using Q-sepharose (Pharmacia Chemical) followed by Sephadex G-75 (Pharmacia Chemical) and HPLC (Mono-Q) chromatography. The purity of the PIase C was checked by SDS-PAGE [20] and showed a single band corresponding to approx. 33 kDa (unpublished data). Surface labelling of platelets with periodate/[3H]boro hydride and their treatments with Plase C Washed platelets suspended in PBS containing 2 mM EDTA were incubated with 2 mM sodium metaperiodate (Sigma Chemical) for 10 rain in the cold (4°C). Platelets were subsequently washed thoroughly with PBS-EDTA buffer by centrifugation and were incubated for 45 rain with 0.5 mCi of NaB3H4 in PBSEDTA buffer [21,22]. Cells were again washed and resuspended in PBS-EDTA or 0.25 M sucrose-EDTA buffer for treatments with BT- or SA-PIase C, respectively. Intact surface-labelled platelets were treated with BT- or SA-Plase C at 37°C for desired time periods and were briefly centrifuged (12 000 × g for 1 min), the resultant supernatant and pellets were separated. The pellet was retained for further use. The supernatant was recentrifuged at 140000 × g for 30 rain. Supernatant was dialyzed exhaustively to eliminate salt or sucrose as necessary and was concentrated by blowing N 2 at 50°C. Pellets and the concentrated supernatant were solubilized in a buffer containing 3~ sodium dodecylsulphate (SDS), 5~ 2-mercaptoethanol, 10~ glycerol, 0.0625 M Tris at pH 6.8 [20]. The samples were subjected to
SDS-polyacrylamide slab gel electrophoresis using 7.5% acrylamide under the conditions described by Laemmli [20]. The gel was stained with Coomassie brilliant blue, destained and soaked in Amplify (Amersham Corporation, IL) for 30 rain with agitation. The gel was dried under vacuum at 80°C using a Bio-Rad slab gel dryer (Model 224, Richmond, CA), exposed to Kodak-X-Omat film at - 7 0 ° C and the fluorograph developed after 48 h exposure. The regions of gel corresponding to radioactive bands were cut out and dissolved in 0.1 ml of 60~ (w/v) perchloric acid and 0.2 ml of 30~ (v/v) H202 for 2 h at 90°C [23]. The dissolved material was then allowed to cool and their radioactivity determined using 10 ml of scintillation cocktail.
Phase separation in Triton X-I I 4 solution After surface labelling platelets with [3H]borohydride, 1~ Triton X-114 [24] Tris-buffered saline (TBS) (pH 7.4), was added in the medium containing platelets (1. 109/ml). It was mixed thoroughly and kept on ice for 30-40 min. The suspension was spun at 4°C and the supernatant transferred to a 1.5 ml conical Eppendorf microfuge tube. This tube wa~ then incubated for 5 rain at 37 ° C. The clear solution became cloudy. It was centrifuged at 500 × g for 3 rain and the detergent appeared as an oily droplet in the bottom, while the upper phase contained the aqueous fraction. The upper aqueous fraction was collected, 0.9 ml of TBS was added in the lower phase, mixed well and kept at 4 °C (ice) for 30 rain. It was again phase-separated by incubation at 37°C followed by centrifugation. The upper phase and the lower detergent phases were collected and concentrated under N2 at 50°C. These samples were solubilized in SDS-cocktail, subjected to SDS-PAGE for separation of proteins and processed for fluorography. After 48-72 h exposure on Kodak X-ray film it was developed to monitor protein patterns. Electrophoretic immunoblotting method After SDS-PAGE of samples, proteins were electrophoreticaUy transferred onto nitrocellulose sheets (BioRad, California) [25]. The electrode buffer contained 25 mM Tris, 192 mM glycine, methanol 20~ (v/v) at pH 8.3. The transfer to nitrocellulose paper was done at 100 V (360 mA) for 5-6 h in cooling conditions. The electrophoretic blots were soaked with 3~ gelatin in Tris-saline (0.9~ NaCI/10 mM Tris-HCl (pH 7.4)) buffer (TBS) for 1 h at room temperature to saturate additional binding sites. The sheets were washed with TBS containing 0.05~ Tween 20 (pH 7.4) and were incubated with monoclonal antibody (mAB) of human platelet glycoprotein lb (GPIb), epitope SZ2 [26], followed by incubation with goat antimouse IgG horseradish peroxidase conjugate (Bio-Rad). The color of the blot was developed using HRP color developing reagent (Bio-Rad) and a photograph was taken immediately.
17
Isolation of platelet membranes Human platelets were separated as described earlier and were suspended in phosphate (10 mM)-buffered saline (155 raM) containing 2 mM EDTA, 100 #M PMSF and 20 #M leupeptin. Platelets were sonicated using a Vibro cell sonicator (Sonics and Materials, Danbury, CT) at setting 3 for eight cycles at 15 s intervals at 4 ° C. The suspension was then centrifuged at 4°C at 600 × g for 10 min. The resultant supernatant was further centrifuged at 140000 × g for 1 h at 4°C. The pellet was resuspended in the PBS buffer containing proteinase inhibitors (EDTA, leupeptine and PMSF), was treated with BT-PIase C at 37°C and processed by procedures described earlier. The samples were subjected to SDS-PAGE and Western blot analysis with mAB of GPIb as described above. Results and Discussion
Fluorography of surface-labelled platelets treated with SA-Plase C or BT-Plase C Human platelets were treated with SA-PIase C for various time periods (0, 15 and 30 rain). The supernatant and pellet fractions (see Materials and Methods) were analyzed by fluorography. Examination of the pattern showed that there is a significant release of a glycoprotein of Mr approx. 150 kDa in a time- and dose-dependent manner (data not shown). The release of the protein by SA-PIase C (5 U / m l ) was maximum during a 30 min time period (Table I). Interestingly, the release of this protein was also observed in the control (i.e., no PIase C) which contained 0.25 M sucrose medium. This background release was less than that after PIase C treatment. It was necessary to use a sucrose medium since SA-PIase C is inhibited by salts and ions [18]. On the basis of this observation an increasing concentration of NaCI was added to a de-
creasing concentration of sucrose to find out whether the release of this protein was due to the depletion of NaCl in the medium. Incubation of surface-labelled platelets with NaC1 for 30 rain did not release this protein significantly but a gradual increase in the concentration of sucrose in the medium with a concomitant decrease in NaCl caused significant release of the 150 kDa glycoprotein. Sucrose (0.25 M) alone released a significant amount of this glycoprotein in 30 min which was about 15% over 0 min control (results not shown). It was also observed that 0.25 M sucrose EDTA buffer and SA-PIase C did not alter the thrombin-stimulated aggregation response [7] or [3H]serotonin secretion (results not shown). These results indicate that depletion of sodium in the medium caused release of a glycoprotein which was also released by SA-PIase C. Therefore, we next used BT-PIase C which is known to be less sensitive to inhibition by sodium chloride than SA-PIase C [27]. Surface-labelled platelets were treated with BT-PIase C (0.5 U/ml) in PBS-EDTA and the release of proteins was monitored using the protocol described in the Methods. In this case also the same 150 kDa glycoprotein was released into the medium after treatment with BT-PIase C (Fig. 1). The release of this glycoprotein by BT-PIase C was compared with SAPIase C and observed that BT-PIase C released a highly significant amount of 150 kDa glycoprotein (P < 0.001) in the medium, whereas SA-PIase C released 10-15% ove~ the control (Table I). BT-PIase C also released this protein more at a concentration of 0.5 U / m l (23%) than 0.25 U (16.0%) or 0.10 U (13.0%) which suggested that the release of this protein from platelet surface by BT-PIase was in a dose-dependent manner (data not shown). The other areas of the gel were cut and counted for radioactivity and no significant release of other components was observed by PIase C treatment (data not shown). This indicates that PIase C treatment re-
TABLE I
Release of glycoprotein by Plase C isolated from Staphylococcus aureus (SA) or Bacillus thuringiensis (BT) Washed human platelets were surface-labelled by sodium borohydride procedure (see Materials and Methods), then treated with BT or SA-PIase C at 37 o C for 30 min. Control platelets were treated with PBS-EDTA buffer (pH 7.2). The platelets were centrifuged and supematant and pellet were separated (see Materials and Methods for details) and were further subjected to SDS-PAGE and fluorography. The regions of gel corresponding to radioactive bands were cut out, dissolved in perchloric acid and H202 and radioactivity counted in 10 ml scintillation cocktail (see Materials and Methods). Values are expressed as the mean_+ S.E.M. for three separate experiments. Figures in the parentheses indicate percentages. Radioactivity (cpm) Control
SA-Plase C (5.00 U/ml)
BT-Plase C (0.50 U/ml)
0 min
30 rain
0 min
30 min
0 rain
30 min
Pellet
28642_+3187 (93.9)
28822_+3240 (91.3)
356234-5733 (91.2)
27901+2958 (84.0)
30508_+3254 (90.3)
27326_+3569 (74.4)
Supernatant
1846+ 158 (6.1)
3720+ 432 (8.6) [A]
34144- 479 (8.7)
5311_+ 427 (15.9) in]
3249_+ 298 (9.6)
Significance: A vs. B, P < 9.05, A vs. C, P < 0.001.
8 433 +
(23.6)
It]
572
18
significantly in the aqueous phase after treatment of surface-labelled platelets with BT-PIase C for 30 rain with a concomitant decrease of radioactivity in the detergent phase. Although the radioactivity in the aqueous phase of control was high (61~), the aqueous phase from PIase C-treated platelets showed radioactivity (87%) which was higher than control. Newman et al. [28] and Clemetson et al. [29] observed that one of the platelet surface glycoprotein, GPIb (150 kDa) partitioned in aqueous phase in Triton X-114 method. Our results of high level of 150 kDa radioactivity in the Triton X-114 aqueous phase in control confirm their observations. Physiological stimuli like thrombin (1 U/ml) and PAF (10 -7 M) slightly increased the radioactivity in the aqueous phase with a concomitant decrease in radioactivity in the detergent phase in comparison to control platelets (Table If). Phase separation in the Triton X-114 solution experiment demonstrated that a high amount of 150 kDa protein partitioned in the aqueous phase without any treatment (i.e., in control). It was earlier observed by Newman et al. [28] and Clemetson et al. [29] that GPIb is found in the Triton X-114-aqueous phase and this may be due to the very high sugar content of GPlb [30,31]. Although the additional appearance of radioactive proteins in aqueous fractions (26% over control) was observed after PIase C treatment, this method has limitations due to partition!ng of most of the 150 kDa protein in the aqueous phase. Thrombin and PAF treatment also showed a slight increase in 150 kDa [aH]glycoprotein in the aqueous phase and these might represent protein released after PI cleavage caused by activation of endogenous Plase C by these stimuli.
150K
Fig, 1, Release of [JHlglycoprotein by BT-Plase C treatment of surface-labelled human platelets. [3H]Borohydride-labelled platelets were treated with BT-Plase C (0.5 U/ml) for 30 min at 37°C and pellets and supernatants were subjected to SDS-PAGE followed by fluorography (see Materials and Methods). (A) 0 min control pellet: (B) 0 min control supernatant; (C) 30 min control pellet: (D) 30 min control supematant; (E) 0 min Plase-treated pellet: (F) 0 min PlaseC-treated supernatant; (O) 30 min Plase C-treated pellet: and (H) 30 rain Plase C-treated supernatant.
leased specifically 150 kDa protein from platelet surface under the experimental conditions described above.
Phase separation with Triton X-I I4 A solution of nonionic detergent Triton X-l14 is homogeneous at 0°C but it separates in an aqueous phase and detergent phase at 20°C or above. This has provided a novel method-for partitioning hydrophilic and hydrophobic proteins [24]. Table II demonstrated that radioactivity of 150 kDa protein was increased
Western blot analysis probed by a monoclonal antibody of GPIb, epitope SZ2 Platelet membrane glycoprotein GPIb is most prominently labelled by the [3H]borohydride procedure used in this study and it also migrates in the same region
TABLE !1 Phase separation in Triton X-114 solution
Surface glycoprotein-labelled human platelets were treated with 1% Triton X-l14 and the detergent and aqueous phases were separated (see Materials and Methods). The other conditions of this experiment were same as in Table 1. In some experiments before, Triton X-114 phase separation platelets were treated with thrombin (IU) or PAF (10 -7) for 1 min. Values are expressed as the mean 4- S.E.M. from two separate experiments. Figures in the parentheses indicate percentages. Radioactivity (cpm) of approx. 150 kDa protein 30 rain
1 rain
control
BT-Plase C
control
thrombin (1 U)
PAF (10- 7)
Aqueous
5 822 4- 632 (60.9) [A]
9133 4-1045 (87.0) [B]
5 969 4- 617 (73.1) [CI
6 855 4- 599 (82.9) [DI
7 752 4- 666 (81.8) [E]
Detergent
3 723 + 423 (39.0)
1364 + 126 (12.9)
2191 + 238 (26.8)
1406 + 135 (17.0)
1715 + 194 (18.1)
Significance: A vs. B, P < 0.05: C vs. D, P < 0.05; C vs. E, P < 0.05.
19 (approx. 150 kDa) [21,22]. We, therefore, examined the reactivity of GPIb mAB to the released surface glycoprotein. Platelets were treated with BT-PIase C (0.5 U,/ml) for 30 min and then centrifuged (see Materials and Methods). The pellet and supernatant of platelets treated with PIase C or without PIase C (control) were solubilized and separated by SDS-polyacrylamide gel electrophoresis. Polypeptides were transferred from gel to nitrocellulose sheets and the sheets were treated with 1 : 1000 dilution of mAB human platelet GPIb, epitope SZ2. The color of the blot was developed by incubation with goat antimouse horseradish peroxidase followed by color reaction (see Materials and Methods). The photograph was taken immediately. Examination of the blot (data not shown) showed that PIase C-treated supernatant reacted strongly with mAB of GPIb in comparison to control. Control supernatant showed little reactivity with GPIb mAB, epitope SZ2. This method was repeated without mAB of GPIb and the reaction of nitrocellulose sheet was negative (results not shown) which ruled out the nonspecific reaction of the released protein. Similar results (Fig. 2) were also observed when the supernatant was obtained from platelets treated with BT-PIase C (0.5 U/ml) in a buffer medium containing proteinase inhibitors EDTA (2 mM), phenyl methyl sulfonyl fluoride [PMSF] (100/~M) and leupeptin (20 /~M). These agents inhibit a wide variety of proteases, including calcium or metal-dependent proteinases (e.g., calpain) and serine and/or thiol proteinases. The positive reactivity with the monoclonal antibody of GPIb, epitope SZ2 suggested that released protein by PIase C treatment represented GPIb complex of platelet surface glycoprotein. It was also established that the glycoprotein released by PIase C action on platelets was not due to the activation of proteinases and that the PIase C activity is unaffected by proteinase inhibitors. Treatment of human platelet membranes with BT-Plase C We next investigated whether the released protein from platelet surface by PIase C is a membrane component or cytosolic contaminant or any proteolytic product of platelet protein. We therefore isolated platelet membranes and treated with BT-PIase C in the presence of proteinase inhibitors (see Methods) and the released surface glycoproteins were examined by immunoreactivity of GPIb mAB, epitope SZ2. It was observed that PIase C-treated membrane fraction supernatant was highly positive, whereas control membrane fraction supernatant showed no reactivity with monoclonal antibody to GPIb (Fig. 3). This observation suggested that approx. 150 kDa component is directly released from the membrane surface of human platelets due to PIase C treatment. In experiments where platelets membranes were used, control supematant showed a band w~ich had faint reactivity to mAB to GPIb and
A
B
C
D
150K
Fig. 2. Western blot analysis of BT-PIase C-treated human platelets probed by a monoclonal antibody of GPIb, epitope SZ2. BT-PIase C (0.5 U / m l ) treatment of intact human platelets was conducted in a medium containing proteinase inhibitors, i.e., EDTA, leupeptin and PMSF (see Results). The samples were then subjected to SDS-PAGE and were electrophoretically transferred to a nitrocellulose sheet. The sheet was then treated with mAb of human platelets GPIb, epitope SZ2, followed by incubation with gout antimouse IgG horseradish peroxidase conjugate. The color of the blot was developed and a photograph was taken immediately. (A) Control pellet; (B) Control supernatant; (C) BT-PIase C-treated pellet; (D) BT-Plase C-treated supernatant.
migrated below the level of approx. 150 kDa (Fig. 3). This band was not seen reproducibly and therefore is not significant. The released 150 kDa glycoprotein reacted with a mAB of human platelet surface GPIb and confirms that the released glycoprotein represents GPIb complex of platelet surface. GPIb has been postulated to be receptor for thrombin in human platelets [14J and also to be involved in vWF-mediated adhesion of platelets to the subendothelium [15]. The epitope SZ2 was directed against GPIb complex and inhibited ristocetin induced platelet agglutination and also collagen- and PAF-induced platelet aggregation [26]. At present, physiological significance of this released 150 kDa protein is unknown and remains to be investigated. It has been suggested that platelet storage results in a redistribution of GPIb molecules and this has provided evidence for a large intraplatelet pool of GPIb [32]. However, our study indicated that the source of released protein is not from the internal pool of GPIb because PIase C caused release of this glycoprotein
20 A
B
C
D
15OK
the 150 kDa protein might reside in the membrane in two types of interactions. A fraction of this protein (e.g., 15-20~) might be exclusively attached to the PI, while the remainder might be putative transmembrane in nature or is somehow resistant to Plase C action. If so, the physiological significance of the differential interactions of the same protein with PI in membranes will be an interesting issue to resolve.
I
Acknowledgements We are grateful to cal assistance and to manuscript. S.D.S. is Development Award
Mr. Lance Antle for expert techniMs. Kristin Nelson for typing the the recipient of a Research Career from NIH (DK 01782).
References 1 2 3 4 5
Fig. 3. BT-Plase C treatment of human platelet membranes. Isolated human platelet membranes were treated with BT-Plase C and the pellet and supernatant fractions were obtained (see Materials and Methods). Both fractions were subjected to Western blot analysis as described in Fig. 2. (A) BT-Plase C-treated membrane pellet; (B) BT-Plase C-treated membrane supernatant; (C) Control membrane pellet: (D) Control membrane supernatant.
from isolated membranes of human platelets which are devoid of this pool. It also confirms that the 150 kDa glycoprotein is directly released from the membrane and that it is not a result of cytosolic proteolytic activity or contaminant due to cellular damage after treatment with Plase C, It is important to note that Plase(s) C is highly purified (see Materials and Methods) and has no effect on the release of [3H]serotonin (data not shown) or aggregation of platelets [7]. The effect of Plase C on the release of 150 kDa protein from isolated membrane in the presence of proteinase inhibitors also confirmed that Plase C actions are not the result of proteinase activity. GPlb is a transmembrane protein and how such a protein can be released, by hydrolysis of Pl from outside, remains to be explained. However, it should be pointed out that several other membrane proteins which are putatively transmembrane in nature (e.g., 5' nucleotidase, decay accelerating factor etc.), and interact with cytoskeleton (e.g., decay accelerating factor, N-CAM etc.), are released by Plase C [2,5]. Roberts et al. [33,34] showed that human erythrocyte acetyl cholinesterase anchored to Pl is insensitive to Plase C due to substitution on the inositol ring by plasmanylinositol (palmitoylated on the inositol ring). It may be envisaged that
Shukla, S.D. (1982) Life Sci. 30, 1323-1335. Low, M.G. (1987) Biochem. J. 244, 1-13. Low, M.G. (1989) FASEB J. 3, 1600-1608. Low, M.G. (1989) Biochim. Biophys. Acta 988, 427-454. Ferguson, M.A.J. and Williams, A.F. (1988) Annu. Rev. Biochem. 57, 285-320. 6 Low, M.G. and Saltiel, A.R. (1988) Science 239, 268-275. 7 Shukla, S.D. (1986) Life Sci. 38, 751-755, 8 Selvaraj, P., Rosse, W.F., Silbert, R. and Springer, T.A. (1988) Nature 333, 565-567. 9 Simmons, D. and Seed, B. (1988) Nature 333, 568-570. 10 Huizinga, T.W.J., Van der Schoot, C.E., Jost, C., Klaassen, R., Kleitjer, M., Von dem Borne, A.E.G.K., Roos, D.A. and Tettroo, P.A.T. (1988) Nature 333, 667-669. 11 Low, M.D and Zilversmit, D.B. (1980) Biochemistry 19, 39133918. 12 Mazumdar, R. and Balasubramariam, A.S. (1985) Biochem. Pharmacol. 34, 4109-4115. 13 Abdel-Latif, A.A. (1986) Pharmacol. Rev, 38, 246-272. 14 Okumura, T., Hasitz, M. and Jameison, G.A. (1978) J. Biol. Chem. 253, 3435-3443, 15 Coiler, B,S., Perrschke, E.I,, Scudder, L.E. and Sullivan, C.A. (1983) Blood 61, 99-110. 16 Philips, D.R. and Baughan, A.K. (1983) J. Biol. Chem. 258, 10240-10246. 17 Dhar, A. and Shukla, S.D. (1988) J. Cell. Biol. 107, 55%. 18 Low, M.D. (1981) Methods Enzymol. 71,741-746. 19 Shukla, S,D., Coleman, R. Finean, J.B. and Michell, R.H. (1980) Biochem. J. 187, 277-280. 20 Lameili, U.K. (1970) Nature 227, 680-685. 21 Dhar, A. and Ganguly, P. (1988) Br. J. Haematoi. 70, 71-75. 22 Steiner, B., Clemetson, K.J. and Luscher, E.F. (1983) Thromb. Res. 29, 43-52. 23 BiUoc, F., Heilmann, E., Combrie, R., Boisseau, M.R. and Nurden, A.T. (1987) Biochim. Biophys. Acta 925, 218-225. 24 Bordier, C. (1981) J. Biol. Chem. 256, 1604-1607. 25 Towbin, H., Staehelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354. 26 Ruan, C., Du, X., Xi, X., Castaldi, P.A. and Berndt, M.C. (1987) Blood 69, 570-577. 27 Taguchi, R., Asahi, Y, and Ikezawa, H. (1980) Biochim. Biophys. Acta 619, 48-57. 28 Newman, P.J., Knipp, M.A. and Kahn, R.A. (1982) Thromb. Res. 27, 221-224.
21 29 Clemetson, K.J., Bienz, D., Zahino, M.-L. and Luscher, E.F. (1984) Biochim. B:ophys. Acta 778, 463-469. 30 Okamura, T., Lo~nbart, C and Jamieson, G.A. (1976) J. Biol. Chem. 251, 5950-5955. 31 Judson, P.A., Anstee, D.J. and Clamp, J.R. (1982) Biochem. J. 205, 81-90.
32 Michelson, A.D., Adelman, B., Barnard, M.R., Carroll, E. and Hardin, R.I. (1988) J. Ciin. Invest. 81, 1734-1740. 33 Roberts, W.L., Myher, J.J., Kuksis, A., Low, M.G. and Rosenberry, T.L. (1988) J. Biol. Chem. 263, 18766-18775. 34 Roberts, W.L., Santika,'n, S., Reinhold, V.N. and Rosenberry, T.L. (1988) J. Biol. Chem. 263, 18776-18784.
